O slideshow foi denunciado.

Hands-On with Google’s Machine Learning APIs, 12/3/2017

1

Compartilhar

•
04 de Dez de 2017
• 1 gostou • 363 visualizações
Próximos SlideShares
Cognitive Services: Building Smart Apps with Speech, NLP & Vision
Cognitive Services: Building Smart Apps with Speech, NLP & Vision
Carregando em…3
×
1 de 46
1 de 46

Hands-On with Google’s Machine Learning APIs, 12/3/2017

04 de Dez de 2017
• 1 gostou • 363 visualizações

1

Compartilhar

Baixar para ler offline

Descrição

Come along with me on a survey of a couple of Google's machine learning APIs: Cloud Vision API and Natural Language. Learn how to build and leverage preexisting models without needing to worry about the complexity of selecting a learning algorithm and running training.

Then, for good measure, we explore the mechanisms behind setting up a basic neural network to classify MNIST digits so that you can run other people's TensorFlow models to classify your own examples.

To follow along with the examples, you will want to check out my GitHub repository at https://github.com/mrcity/mlworkshop/ for example code and setup instructions, including how to run an existing TensorFlow model from your own code in order to classify examples.

Transcrição

  1. 1. Hands-On with Google’s Machine Learning APIs Stephen Wylie 12/3/2017 @SWebCEO +StephenWylie mrcity
  2. 2. About me WEBSITES www.stev-o.us goshtastic.blogspot.com www.ledgoes.com www.openbrite.com EMAIL stev-o@stev-o.us nerds@openbrite.com G+ https://plus.google.com/u/1 /+StephenWylie TWITTER @SWebCEO GITHUB mrcity  Senior Software Engineer at Capital One  Skunkworks for Auto Finance Innovation  Successful Kickstarter (1000% funded - BriteBlox)  Intel Innovator, DMS Member, Vintage computer collector/restorer/homebrewer, hackathoner  Civic Hacking Cmte Chair @ DMS @SWebCEO +StephenWylie #MachineLearning
  3. 3. Today’s Mission  Touch on ML API offerings  Explore Google’s RESTful ML APIs  Cloud Vision  Natural Language  TensorFlow (Not a RESTful API but still cool)  Allocate time to play, develop ideas  Have good conversations, network  Find learning partner or group @SWebCEO +StephenWylie #MachineLearning
  4. 4. Before We Start… Hopefully you followed instructions on https://github.com/mrcity/mlworkshop/  Get access to the APIs  Install TensorFlow @SWebCEO +StephenWylie #MachineLearning
  5. 5. What is Machine Learning?  Programming computers to deduce things from data…  Conclusions  Patterns  Objects in images  …using generic mathematical methods  No advance knowledge of trends in data  Lots of algorithms available  The process can create beautiful constructs @SWebCEO +StephenWylie #MachineLearning
  6. 6. Who’s using ML?  Chat bots  Self-driving cars  Pepper the robot  MarI/O  Document recognition & field extraction @SWebCEO +StephenWylie #MachineLearning
  7. 7. Machine Learning Tools Ecosystem  APIs you interface with  HP, Amazon, Microsoft, IBM, Google, Facebook’s Caffe on mobile & Web  Software you use  Orange (U of Ljubljana, Slovenia)  Weka (U of Waikato, New Zealand)  Hardware you compile programs to run on  nVidia GPUs with CUDA, DGX-1 supercomputer  Movidius neural compute stick  Amazon DeepLens camera @SWebCEO +StephenWylie #MachineLearning
  8. 8. Google’s ML APIs In Particular  Google Play Services  Mobile Vision API  RESTful ML API Services  Cloud Vision, Cloud Speech, Natural Language, Translation  Job Discovery*, DialogFlow, Cloud Video Intelligence  Cloud ML Engine  Local ML Services  TensorFlow, Tensorflow Lite  Tensorflow Serving @SWebCEO +StephenWylie #MachineLearning ^ Pre-defined models v User-defined models * private beta
  9. 9. https://cloud.google.com/vision/ Cloud Vision API What does it look like to you? @SWebCEO +StephenWylie #MachineLearning
  10. 10. Detect Faces, Parse Barcodes, Segment Text @SWebCEO +StephenWylie #MachineLearning Availability Native Android Native iOS RESTful API FACE API BARCODE API TEXT API
  11. 11. What do you see in that cloud? Breaks down into more features than just FACE, BARCODE, and TEXT: From https://cloud.google.com/vision/docs/requests-and-responses @SWebCEO +StephenWylie #MachineLearning Feature Type Description LABEL_DETECTION Execute Image Content Analysis on the entire image and return TEXT_DETECTION Perform Optical Character Recognition (OCR) on text within an image FACE_DETECTION Detect faces within the image LANDMARK_DETECTION Detect geographic landmarks within the image LOGO_DETECTION Detect company logos within the image SAFE_SEARCH_DETECTION Determine image safe search properties on the image IMAGE_PROPERTIES Compute image properties, including dominant colors
  12. 12. What do you see in that cloud? Beta API 1.1 offers additional features: Note Google does not guarantee any SLAs, deprecation policies, or future backward compatibility with these services. @SWebCEO +StephenWylie #MachineLearning Feature Type Description IMAGE_PROPERTIES Predict crop hints for the image, in addition to previous data WEB_DETECTION Detect news, events, or celebrities within an image, then search Google Images for similar photos DOCUMENT_TEXT_ DETECTION Optimize text parser for dense OCR, rather than some text with non-text
  13. 13. Cloud Vision APIs  Can simultaneously detect multiple features  Features billed individually per use on image  No Barcode feature  Simple JSON request/response format  Submit image from Cloud Storage or in Base64  Returns 0 or more annotations by confidence @SWebCEO +StephenWylie #MachineLearning
  14. 14. For Your Eyes Only No OAuth required for Cloud Vision Make requests using API Key POST https://vision.googleapis.com/v1/images:annotate?key={YOUR_API_KEY} Easy to script using Service Account @SWebCEO +StephenWylie #MachineLearning
  15. 15. Response types @SWebCEO +StephenWylie #MachineLearning Feature Returns Label Description of the picture’s contents Confidence score Text, Logo Text contents or logo owner name Bounding polygon containing the text or logo [Text only] Locale (language) [Logo only] Confidence score Face Bounding polygon and rotational characteristics of the face Positions of various characteristics such as eyes, ears, lips, chin, forehead Confidence score of exhibiting joy, sorrow, anger, or surprise Landmark Description of the landmark and confidence score Bounding polygon of the recognized landmark in the picture Safe Search Likelihood of the image containing adult or violent content, that it was a spoof, or contains graphic medical imagery
  16. 16. Response types @SWebCEO +StephenWylie #MachineLearning Feature Returns Image properties Array of dominant RGB colors within the image, ordered by fraction of pixels Crop hints by confidence and “importanceFraction” Web Entities and Pages News, events, celebrities, or other labels found within an image Similar images from Google Image Search Website URLs containing matching images Document Text List of nested objects consisting of types Page  Block  Paragraph  Word  Symbol Bounding box of the recognized text Good example of the Document Text hierarchy at https://cloud.google.com/vision/docs/detecting-fulltext#vision-document- text-detection-python
  17. 17. Demo @SWebCEO +StephenWylie #MachineLearning
  18. 18. Mobile Vision vs. Cloud Vision Mobile Vision is for Native Android Free; no usage quotas Handles more data processing Can utilize camera video Takes advantage of hardware @SWebCEO +StephenWylie #MachineLearning
  19. 19. https://cloud.google.com/natural-language/ Cloud Natural Language API Making computers speak human @SWebCEO +StephenWylie #MachineLearning
  20. 20. Natural Language API: Analyze Any ASCII Parses text for parts of speech Discovers entities like organizations, people, locations Analyzes text sentiment Use Speech, Vision, Translate APIs upstream Works with English, Spanish, or Japanese @SWebCEO +StephenWylie #MachineLearning
  21. 21. Sample NL API Request @SWebCEO +StephenWylie #MachineLearning From https://cloud.google.com/natural-language/docs/basics <- Optional, can be guessed automatically <- Optional, but recommended // Use only ONE of “content” or “gcsContentUri” Pick one of https://language.googleapis.com/v1/documents:analyzeEntities analyzeEntitySentiment analyzeSentiment analyzeSyntax classifyText
  22. 22. Combined Sample NL API Request @SWebCEO +StephenWylie #MachineLearning From https://cloud.google.com/natural-language/docs/basics <- Optional, can be guessed automatically Make request to https://language.googleapis.com/v1/documents:annotateText
  23. 23. Interpreting NL API Sentiment Responses @SWebCEO +StephenWylie #MachineLearning SCORE -1 1 MAGNITUDE 0 ∞ 1 10 10 2 10 3 Sample analyzeSentiment response for the Gettysburg Address: { “score”: 0.4, “magnitude”: 3.8 }
  24. 24. Demo @SWebCEO +StephenWylie #MachineLearning
  25. 25. https://cloud.google.com/prediction/ Google Prediction API (Deprecated) To further their conquest for all knowledge past, present, and future This service is shutting down on 4/30/2018. @SWebCEO +StephenWylie #MachineLearning
  26. 26. Making Predictions With Google  Build “trained” model or use “hosted” model  Hosted models (all demos):  Language identifier  Tag categorizer (as android, appengine, chrome, youtube)  Sentiment predictor  Trained models:  Submit attributes and labels for each example  Need at least six examples  Store examples in Cloud Storage @SWebCEO +StephenWylie #MachineLearning
  27. 27. Don’t Model Trains; Train Your Model Train API against data prediction.trainedmodels.insert Send prediction query prediction.trainedmodels.predict Update the model prediction.trainedmodels.update Other CRUD operations: list, get, delete @SWebCEO +StephenWylie #MachineLearning
  28. 28. Don’t Model Trains; Train Your Model Insert query requires: id modelType storageDataLocation Don’t forget: poll for status updates @SWebCEO +StephenWylie #MachineLearning
  29. 29. Permissions To Make Predictions OAuth is required for Predictions Easy to script using Service Account Or, get Web app credentials: https://console.developers.google.com/apis/credentials @SWebCEO +StephenWylie #MachineLearning
  30. 30. Demo @SWebCEO +StephenWylie #MachineLearning
  31. 31. Smartly Auto Fill Google Sheets Add-on for Google Spreadsheets No previously trained model needed Use data with partially-labeled examples Specify column to auto-fill Wait... a... long... time... @SWebCEO +StephenWylie #MachineLearning
  32. 32. Demo @SWebCEO +StephenWylie #MachineLearning
  33. 33. TensorFlow All that Linear Algebra you slept through in college @SWebCEO +StephenWylie #MachineLearning
  34. 34. About TensorFlow  Offline library for large- scale numerical computation  Think of a graph:  Nodes represent mathematical operations  Edges represent tensors flowing between them  Excellent at building deep neural networks @SWebCEO +StephenWylie #MachineLearning Soft𝑚𝑎𝑥 𝑥 𝑖 = 𝑒 𝑥 𝑖 𝑗 𝑒 𝑥 𝑗 𝑅𝑒𝐿𝑈 𝑛 =𝑓 𝑥 = max(0, 𝑥)
  35. 35. Tense About Tensors?  Think about MNIST handwritten digits Each number is 28 pixels squared There are 10 numbers, 0-9 @SWebCEO +StephenWylie #MachineLearning
  36. 36. Tense About Tensors?  Define an input tensor of shape (any batch size, 784) x = tf.placeholder(tf.float32, shape=[None, 784])  Define a target output tensor of shape (any batch size, 10) y_ = tf.placeholder(tf.float32, shape=[None, 10])  Define weights matrix (784x10) and biases vector (10-D) @SWebCEO +StephenWylie #MachineLearning
  37. 37. One-Hot: Cool To the Touch  Load the input data from tensorflow.examples.tutorials.mnist import input_data mnist = input_data.read_data_sets('MNIST_data', one_hot=True)  One-hot?!  Think about encoding categorical features: US = 0, UK = 1, India = 2, Canada = 3, …  This implies ordinal properties and confuses learners  Break encoding into Booleans:  This is where the 10-D target output tensor comes from @SWebCEO +StephenWylie #MachineLearning US = [1, 0, 0, 0] UK = [0, 1, 0, 0] Etc…
  38. 38. TensorFlow Data Structures - Placeholders  Come from inputs prior to computation  x (input picture as vector), y_ (one-hot 10-D classification vector) @SWebCEO +StephenWylie #MachineLearning [0, 0, 0, 0, 0, 0, 0, 0, … 0, 0, 0, 0, 0, 1, 1, 0, … 0, 0, 0, 0, 1, 1, 1, 0, … 0, 0, 0, 0, 1, 1, 0, 0, … 0, 0, 0, 1, 1, 1, 0, 0, … ... [0, 0, 0, 0, 1, 0, 0, 0, 0, 0] Input x y_
  39. 39. TensorFlow Data Structures – Variables  Values (i.e. model parameters) inside nodes  Used and modified by learning process  Need to be initialized with W = tf.Variable(tf.zeros([784,10])) init = tf.global_variables_initializer()  W (weights to scale inputs by), b (bias to add to scaled value) @SWebCEO +StephenWylie #MachineLearning
  40. 40. Training a Dragon, if the Dragon is a Model  Your Simple Model: y = tf.matmul(x, W) + b  Cross-entropy: distance between guess & correct answer cross_entropy = tf.reduce_mean(tf.nn. softmax_cross_entropy_with_logits(labels=y_, logits=y))  Gradient descent: minimize cross-entropy train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)  Learning rate: 0.5 @SWebCEO +StephenWylie #MachineLearning 𝐻 𝑦′ 𝑦 = − 𝑖 𝑦𝑖 ′ log(𝑦𝑖)
  41. 41. Dragon Get Wiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiings!  Start a Session  Run global_variables_initializer  Run training for 1000 steps sess = tf.Session() sess.run(init) for i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) train_step.run(feed_dict={x: batch_xs, y_: batch_ys})  Expensive to use all training data at once!  Pick 100 random samples each step @SWebCEO +StephenWylie #MachineLearning
  42. 42. Test Flight Evaluation  Compare labels between guess y and correct y_ correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))  Cast each Boolean result into either a 0 or 1, then average it accuracy = tf.reduce_mean( tf.cast(correct_prediction, tf.float32))  Print the final figure print(accuracy.eval(feed_dict={ x: mnist.test.images, y_: mnist.test.labels})) @SWebCEO +StephenWylie #MachineLearning
  43. 43. Demo @SWebCEO +StephenWylie #MachineLearning
  44. 44. Future Talks, Other Talks Follow me if you want to hear these!  Build a Neural Network in Python with NumPy  Build a Neural Network with nVidia CUDA Elsewhere,  Mapmaking with Google Maps API, Polymer, and Firebase  The Process Of Arcade Game ROM Hacking @SWebCEO +StephenWylie #MachineLearning
  45. 45. More Resources  Google’s “Googly Eyes” Android app [Mobile Vision API] https://github.com/googlesamples/android- vision/tree/master/visionSamples/googly-eyes  Quick, Draw! Google classification API for sketches https://quickdraw.withgoogle.com/  Making Android Apps With Intelligence, by Margaret Maynard- Reid https://realm.io/news/360andev-margaret-maynard-reid- making-android-apps-with-intelligence/ (Video + slides) @SWebCEO +StephenWylie #MachineLearning
  46. 46. Thank You @SWebCEO +StephenWylie #MachineLearning

Notas do Editor

  • Some machine learning algorithms work better for certain types of problems than others. However, basically all of them start out with a blank slate, or perhaps random data as intermediate values, and then the dataset is analyzed in order to find which attributes of an example outcome provide the most information gain in order to make the correct classification on the training data. Learning happens iteratively over time in a manner that tries to come up with an ever more accurate answer.
  • Chat bots need machine learning in order to parse the conversation into its grammatical components, and to understand the context of the conversation so it can come up with an appropriate response.

    Pepper the robot is used for a variety of emotion-based tasks such as teaching, helping kids with autism stay focused, and training people to make standard responses to events if they are not socially adjusted.

    I hate to say this, because my mom works in Accounts Payable, but eventually her job will be automated by a system that can automatically scan an invoice, look for the amount due, and facilitate the payment over a payment network electronically. Document recognition is important for banks too, as statements which prove income for purposes of granting a loan come in many different forms.
  • HPE Haven – text analysis, speech recognition, image recognition, face detection, recommendation engine, knowledge graph analysis
    Amazon – industry-standard ML algorithms on top of your own data sets with listeners to re-evaluate upon new data (kind of like the now-deprecated Google Cloud Prediction API)
    Amazon also just rolled out SageMaker, which is a cloud-based service to help users visualize data and select the best ML algorithm, then deploy it at scale for the inference stage
    Amazon Rekognition image & video analysis that ties in with the DeepLens camera
    Azure APIs – language analytics, face & emotion & explicit content detection, speech recognition, and recommendation APIs
    IBM Watson
    Facebook – Ported Caffe for several platforms to run light & efficiently

  • Things above the line are pre-defined models that Google has already tuned for you to give what they feel are the best results. Things below the line are where you have to provide the data and possibly the tuning yourself in order to get the best outcome.

    Tensorflow Serving falls in Local because you have to set up the parameters of the infrastructure yourself, whereas Cloud ML Engine will size the model inference environment for you automatically. However, TF Serving can also be run on any hardware you have, not just inside GCP.
  • The native implementations are brought to you by the Mobile Vision API. The RESTful API which is called Cloud Vision does not feature barcode reading capabilities since that’s something a device should be able to handle on its own natively.
  • Safe search includes Adult, Medical, Spoof, and Violence.
  • No Barcode feature because why should you send images over the network just to find out if it’s a barcode when chances are it’s not.
  • Before we get the response, there’s one more detail we need to form our request: the authentication.

    Google provides several ways to allow use of the Machine Learning APIs, and the exact ways you implement them depend on what level of user data you seek to access in your application.

    The Cloud Vision API does not require you to pass along OAuth prompts to your users. You can simply make a request by finding your browser API key (usually starting with the letters AIza) and then appending that to your query string. However, this precludes you from using Google’s handy dandy API libraries for your programming language of choice. (Then again, if you’re using ALGOL 68, SmallTalk, or 6502 assembly, maybe it’s your only option.)

    To use Google’s libraries, you need to at least set up a Service Account. This type of authentication scheme is usually used for servers communicating to other servers and requires minimal end user intervention. Also, typical controls you might find on accounts in a Google Apps domain do not apply to service accounts, therefore you could be inadvertently granting your users the chance to do something dangerous like share documents and data outside of your domain if you are not careful with the permissions you grant them.

    Nevertheless, to save ourselves some time and trouble, we are going to use the basic Service Account approach for the sake of this demo. The instructions on how to create an appropriate service worker are on this workshop’s GitHub repo.
  • importanceFraction is the “fraction of importance of this salient region with respect to the original image.” Not really sure how they come to that conclusion.
  • This demo consists of running the Googly Eyes Node.js application provided in my GitHub repo. With some files in my Google Cloud Storage bucket, I could refer to that storage bucket name and the file name to load up files with human faces in them, and then this program would use the parameters returned by the Face Detection API and do some mathematics to superimpose googly eyes on top of the image using HTML5 Canvas.
  • One would presume the native libraries would be taking advantage of lower-level calls to get to the bare metal of the phone and make the analysis faster.
  • In the past, Sentiment Analysis was only available for English text. The Changelog https://cloud.google.com/natural-language/release-notes from 11/15/16 indicates that Japanese & Spanish support is available on sentiment analysis, but their How-To guide for “Analyzing Sentiment” indicates only English is supported. What’s interesting is when an English string is scored, then sent through Translate, and then you see totally different sentiment scores.
  • There are client libraries for C#, Go, Java, and other languages, plus RPC calls, available for interaction with the NL API. For RESTful calls, this is the syntax to use if you wish to just run one particular NL API query at a time.

    analyzeEntities gives you phrases in the text that are known entities, such as persons, locations, or organizations, and references the saliency of the entity to the article as well as how many times it was mentioned.
    analyzeSentiment gives you sentiment scores for the entire text body, and analyzeEntitySentiment will give you sentiment scores for text associated with each entity and its mentions. More on this in the next slide.
    analyzeSyntax gives you all the parts of speech and grammatical considerations for each word.
    classifyText will give you an array of potential categories for the text as well as the confidence of that category classification.
  • In this construct with the “annotateText” endpoint invoked, specify in the “features” object which of the three types of NLP queries you want to run.
  • The score (formerly known as polarity) is a [-1, 1] scale of the emotional sentiment (tone) of the article, from negative/scathing/blistering to positive/gushing/cheerful.

    The magnitude of the article is a scale of how emotional the document comes across. Since each expression in the text contributes to the magnitude, longer articles will have a higher magnitude, regardless of whether the tone of each individual emotional word in the text is positive or negative in sentiment.

    The small snippet of JSON here is now contained within an object called “documentSentiment”. A sibling of this object in the returned JSON structure is called “sentences”, which is an array consisting of objects containing the text and offset of each sentence, plus a “sentiment” object containing the individual magnitude and score for that piece.
  • In this demo, I simply navigated to the Web sandbox at https://cloud.google.com/natural-language/ to run various strings of text through the tool.
  • Types of models:
    - Regression: estimate a numeric answer based on the examples along a continuous curve, akin to a formula that you would solve for
    - Categorization: group possible answers into buckets that don’t have a clear distance to other buckets
  • Here again, we need to discuss authentication, because it’s a little bit stricter for the Prediction API than for Cloud Vision.

    Users must be authenticated into the Prediction API in order to use it. We can still use Service Accounts to authenticate ourselves into the app, but you have to specify a few more settings when creating them, as specified in this workshop’s GitHub repository.

    However, to push the authentication screen to your end users, go into the API Manager (at console.*developers*.google.com), open up the Credentials page, and create an OAuth Client ID credential. It will guide you through steps based on the type of devices your application is targeting, and then use the client secrets file it creates when you are defining the client object in your application. Depending on the language you are using, the objects you create to authenticate a user through a Web-based service or a standalone application might look different than that used for Service Accounts.
  • In this example, I took the data set from Problem Set #2 from my old Machine Learning homework from Fall 2010 (at http://users.eecs.northwestern.edu/~ahu340/eecs349-ps2/ ), and loaded in the train.csv file into my Google Cloud Storage Bucket. I had to massage the CSV file into the exact format it was looking for. This includes things like removing all double quotes and changing the outcome of the example to a text value rather than a numeric value so it would classify the outcome as a categorization rather than a regression, since we probably don’t want to force the values of “0” for no and “1” for yes to be the regression we solve for.

    After loading in the training data, I took some of the classification examples and pasted them into the Prediction API sample app running locally, downloadable from this workshop’s GitHub repo.

    If you want to try this in API Explorer, break out the attributes of each example into JSON in the following format (note if copying the following to watch for unintended fancy double quotes):

    {
    "input": {
    "csvInstance": [
    "104042",
    "0",
    "77",
    "m",
    "s",
    "0",
    "9",
    "7",
    "523",
    "428824",
    "24",
    "115315"
    ]
    }
    }
  • In this example, I took the data set from Problem Set #2 from my old Machine Learning homework from Fall 2010 (at http://users.eecs.northwestern.edu/~ahu340/eecs349-ps2/ ), and loaded in the train.csv file into my Google Cloud Storage Bucket. I had to massage the CSV file into the exact format it was looking for. This includes things like removing all double quotes and changing the outcome of the example to a text value rather than a numeric value so it would classify the outcome as a categorization rather than a regression, since we probably don’t want to force the values of “0” for no and “1” for yes to be the regression we solve for.

    After loading in the training data, I took some of the classification examples and pasted them into the Prediction API sample app running locally, downloadable from this workshop’s GitHub repo.

    If you want to try this in API Explorer, break out the attributes of each example into JSON in the following format (note if copying the following to watch for unintended fancy double quotes):

    {
    "input": {
    "csvInstance": [
    "104042",
    "0",
    "77",
    "m",
    "s",
    "0",
    "9",
    "7",
    "523",
    "428824",
    "24",
    "115315"
    ]
    }
    }
  • The ReLU (rectified linear unit) function is a popular function to use inside perceptrons.

    We will describe the Softmax function in great detail later.
  • The old way, initialize_all_variables(), should have been removed back in March 2017.
  • The softmax function is one that attempts to normalize the vector of guesses so that all the outputs sum to exactly 1, thus converting the output of guesses into a probability distribution. It also serves to multiplicatively increase the weight given to any particular outcome per additional unit of evidence, but also reduce that weight with each unit of evidence withheld. Thus, it is often used to represent categorical distributions. After all, we are categorizing pictures of digits as 0 through 9; that doesn’t mean they will exist linearly in space away from each other, or even in the same order, if you were to look at the values of each attribute that makes the system think it’s one particular digit compared to another.

    Cross-entropy is used as a function to compute the cost of encoding a particular event drawn from a set. The less information required to make the distinction, the better. In this case, we can also represent how far apart the guess is from the correct answer on the training set. Cross-entropy helps perform learning faster by avoiding slowdowns that could be encountered by other functions like the quadratic cost function, where learning is impaired when it gets close to the correct answer.

    Incidentally, the formula for cross-entropy has changed slightly and has been updated in this presentation. It was noted in the comments of the example code that the previous way could become numerically unstable, and so it has been changed to what is listed here. This also simplifies the model because the new function will handle taking the softmax of y.
  • The first code snippet takes the vectors and compares the first dimension of each element between y and y_. It then generates a new vector equal to the length of y (or y_ for that matter) and each result is a Boolean as to whether the things being compared were equal or not.

    The second code snippet converts each Boolean into a 0 or 1, and then takes the average of all the elements in the vector to give you accuracy as a fraction.

    The third code snippet actually causes the accuracy to be calculated. Note this used to be print(sess.run(accuracy, feed_dict=…)) and it’s still like that in the code examples, but the way listed here is in the latest documentation and it works fine.
  • In this demo, I launchTensorFlow through a Docker instance and ran my Python code that runs a previously-generated model to classify one of the images represented in one of the datasets.

    If you happened to terminate a previously-running Docker instance, you can reactivate it and get into it again with the following commands on the command line:
    ed
    docker start `docker ps -q -l`
    docker attach `docker ps -q -l`
  • Note there are several versions of Margaret’s talk available to be found on the Internet, including one from Big Android BBQ 2016 that’s a couple months newer and might have slightly different content (like how this slide deck is different from November’s deck).
    • Software

    Descrição

    Come along with me on a survey of a couple of Google's machine learning APIs: Cloud Vision API and Natural Language. Learn how to build and leverage preexisting models without needing to worry about the complexity of selecting a learning algorithm and running training.

    Then, for good measure, we explore the mechanisms behind setting up a basic neural network to classify MNIST digits so that you can run other people's TensorFlow models to classify your own examples.

    To follow along with the examples, you will want to check out my GitHub repository at https://github.com/mrcity/mlworkshop/ for example code and setup instructions, including how to run an existing TensorFlow model from your own code in order to classify examples.

    Transcrição

    1. 1. Hands-On with Google’s Machine Learning APIs Stephen Wylie 12/3/2017 @SWebCEO +StephenWylie mrcity
    2. 2. About me WEBSITES www.stev-o.us goshtastic.blogspot.com www.ledgoes.com www.openbrite.com EMAIL stev-o@stev-o.us nerds@openbrite.com G+ https://plus.google.com/u/1 /+StephenWylie TWITTER @SWebCEO GITHUB mrcity  Senior Software Engineer at Capital One  Skunkworks for Auto Finance Innovation  Successful Kickstarter (1000% funded - BriteBlox)  Intel Innovator, DMS Member, Vintage computer collector/restorer/homebrewer, hackathoner  Civic Hacking Cmte Chair @ DMS @SWebCEO +StephenWylie #MachineLearning
    3. 3. Today’s Mission  Touch on ML API offerings  Explore Google’s RESTful ML APIs  Cloud Vision  Natural Language  TensorFlow (Not a RESTful API but still cool)  Allocate time to play, develop ideas  Have good conversations, network  Find learning partner or group @SWebCEO +StephenWylie #MachineLearning
    4. 4. Before We Start… Hopefully you followed instructions on https://github.com/mrcity/mlworkshop/  Get access to the APIs  Install TensorFlow @SWebCEO +StephenWylie #MachineLearning
    5. 5. What is Machine Learning?  Programming computers to deduce things from data…  Conclusions  Patterns  Objects in images  …using generic mathematical methods  No advance knowledge of trends in data  Lots of algorithms available  The process can create beautiful constructs @SWebCEO +StephenWylie #MachineLearning
    6. 6. Who’s using ML?  Chat bots  Self-driving cars  Pepper the robot  MarI/O  Document recognition & field extraction @SWebCEO +StephenWylie #MachineLearning
    7. 7. Machine Learning Tools Ecosystem  APIs you interface with  HP, Amazon, Microsoft, IBM, Google, Facebook’s Caffe on mobile & Web  Software you use  Orange (U of Ljubljana, Slovenia)  Weka (U of Waikato, New Zealand)  Hardware you compile programs to run on  nVidia GPUs with CUDA, DGX-1 supercomputer  Movidius neural compute stick  Amazon DeepLens camera @SWebCEO +StephenWylie #MachineLearning
    8. 8. Google’s ML APIs In Particular  Google Play Services  Mobile Vision API  RESTful ML API Services  Cloud Vision, Cloud Speech, Natural Language, Translation  Job Discovery*, DialogFlow, Cloud Video Intelligence  Cloud ML Engine  Local ML Services  TensorFlow, Tensorflow Lite  Tensorflow Serving @SWebCEO +StephenWylie #MachineLearning ^ Pre-defined models v User-defined models * private beta
    9. 9. https://cloud.google.com/vision/ Cloud Vision API What does it look like to you? @SWebCEO +StephenWylie #MachineLearning
    10. 10. Detect Faces, Parse Barcodes, Segment Text @SWebCEO +StephenWylie #MachineLearning Availability Native Android Native iOS RESTful API FACE API BARCODE API TEXT API
    11. 11. What do you see in that cloud? Breaks down into more features than just FACE, BARCODE, and TEXT: From https://cloud.google.com/vision/docs/requests-and-responses @SWebCEO +StephenWylie #MachineLearning Feature Type Description LABEL_DETECTION Execute Image Content Analysis on the entire image and return TEXT_DETECTION Perform Optical Character Recognition (OCR) on text within an image FACE_DETECTION Detect faces within the image LANDMARK_DETECTION Detect geographic landmarks within the image LOGO_DETECTION Detect company logos within the image SAFE_SEARCH_DETECTION Determine image safe search properties on the image IMAGE_PROPERTIES Compute image properties, including dominant colors
    12. 12. What do you see in that cloud? Beta API 1.1 offers additional features: Note Google does not guarantee any SLAs, deprecation policies, or future backward compatibility with these services. @SWebCEO +StephenWylie #MachineLearning Feature Type Description IMAGE_PROPERTIES Predict crop hints for the image, in addition to previous data WEB_DETECTION Detect news, events, or celebrities within an image, then search Google Images for similar photos DOCUMENT_TEXT_ DETECTION Optimize text parser for dense OCR, rather than some text with non-text
    13. 13. Cloud Vision APIs  Can simultaneously detect multiple features  Features billed individually per use on image  No Barcode feature  Simple JSON request/response format  Submit image from Cloud Storage or in Base64  Returns 0 or more annotations by confidence @SWebCEO +StephenWylie #MachineLearning
    14. 14. For Your Eyes Only No OAuth required for Cloud Vision Make requests using API Key POST https://vision.googleapis.com/v1/images:annotate?key={YOUR_API_KEY} Easy to script using Service Account @SWebCEO +StephenWylie #MachineLearning
    15. 15. Response types @SWebCEO +StephenWylie #MachineLearning Feature Returns Label Description of the picture’s contents Confidence score Text, Logo Text contents or logo owner name Bounding polygon containing the text or logo [Text only] Locale (language) [Logo only] Confidence score Face Bounding polygon and rotational characteristics of the face Positions of various characteristics such as eyes, ears, lips, chin, forehead Confidence score of exhibiting joy, sorrow, anger, or surprise Landmark Description of the landmark and confidence score Bounding polygon of the recognized landmark in the picture Safe Search Likelihood of the image containing adult or violent content, that it was a spoof, or contains graphic medical imagery
    16. 16. Response types @SWebCEO +StephenWylie #MachineLearning Feature Returns Image properties Array of dominant RGB colors within the image, ordered by fraction of pixels Crop hints by confidence and “importanceFraction” Web Entities and Pages News, events, celebrities, or other labels found within an image Similar images from Google Image Search Website URLs containing matching images Document Text List of nested objects consisting of types Page  Block  Paragraph  Word  Symbol Bounding box of the recognized text Good example of the Document Text hierarchy at https://cloud.google.com/vision/docs/detecting-fulltext#vision-document- text-detection-python
    17. 17. Demo @SWebCEO +StephenWylie #MachineLearning
    18. 18. Mobile Vision vs. Cloud Vision Mobile Vision is for Native Android Free; no usage quotas Handles more data processing Can utilize camera video Takes advantage of hardware @SWebCEO +StephenWylie #MachineLearning
    19. 19. https://cloud.google.com/natural-language/ Cloud Natural Language API Making computers speak human @SWebCEO +StephenWylie #MachineLearning
    20. 20. Natural Language API: Analyze Any ASCII Parses text for parts of speech Discovers entities like organizations, people, locations Analyzes text sentiment Use Speech, Vision, Translate APIs upstream Works with English, Spanish, or Japanese @SWebCEO +StephenWylie #MachineLearning
    21. 21. Sample NL API Request @SWebCEO +StephenWylie #MachineLearning From https://cloud.google.com/natural-language/docs/basics <- Optional, can be guessed automatically <- Optional, but recommended // Use only ONE of “content” or “gcsContentUri” Pick one of https://language.googleapis.com/v1/documents:analyzeEntities analyzeEntitySentiment analyzeSentiment analyzeSyntax classifyText
    22. 22. Combined Sample NL API Request @SWebCEO +StephenWylie #MachineLearning From https://cloud.google.com/natural-language/docs/basics <- Optional, can be guessed automatically Make request to https://language.googleapis.com/v1/documents:annotateText
    23. 23. Interpreting NL API Sentiment Responses @SWebCEO +StephenWylie #MachineLearning SCORE -1 1 MAGNITUDE 0 ∞ 1 10 10 2 10 3 Sample analyzeSentiment response for the Gettysburg Address: { “score”: 0.4, “magnitude”: 3.8 }
    24. 24. Demo @SWebCEO +StephenWylie #MachineLearning
    25. 25. https://cloud.google.com/prediction/ Google Prediction API (Deprecated) To further their conquest for all knowledge past, present, and future This service is shutting down on 4/30/2018. @SWebCEO +StephenWylie #MachineLearning
    26. 26. Making Predictions With Google  Build “trained” model or use “hosted” model  Hosted models (all demos):  Language identifier  Tag categorizer (as android, appengine, chrome, youtube)  Sentiment predictor  Trained models:  Submit attributes and labels for each example  Need at least six examples  Store examples in Cloud Storage @SWebCEO +StephenWylie #MachineLearning
    27. 27. Don’t Model Trains; Train Your Model Train API against data prediction.trainedmodels.insert Send prediction query prediction.trainedmodels.predict Update the model prediction.trainedmodels.update Other CRUD operations: list, get, delete @SWebCEO +StephenWylie #MachineLearning
    28. 28. Don’t Model Trains; Train Your Model Insert query requires: id modelType storageDataLocation Don’t forget: poll for status updates @SWebCEO +StephenWylie #MachineLearning
    29. 29. Permissions To Make Predictions OAuth is required for Predictions Easy to script using Service Account Or, get Web app credentials: https://console.developers.google.com/apis/credentials @SWebCEO +StephenWylie #MachineLearning
    30. 30. Demo @SWebCEO +StephenWylie #MachineLearning
    31. 31. Smartly Auto Fill Google Sheets Add-on for Google Spreadsheets No previously trained model needed Use data with partially-labeled examples Specify column to auto-fill Wait... a... long... time... @SWebCEO +StephenWylie #MachineLearning
    32. 32. Demo @SWebCEO +StephenWylie #MachineLearning
    33. 33. TensorFlow All that Linear Algebra you slept through in college @SWebCEO +StephenWylie #MachineLearning
    34. 34. About TensorFlow  Offline library for large- scale numerical computation  Think of a graph:  Nodes represent mathematical operations  Edges represent tensors flowing between them  Excellent at building deep neural networks @SWebCEO +StephenWylie #MachineLearning Soft𝑚𝑎𝑥 𝑥 𝑖 = 𝑒 𝑥 𝑖 𝑗 𝑒 𝑥 𝑗 𝑅𝑒𝐿𝑈 𝑛 =𝑓 𝑥 = max(0, 𝑥)
    35. 35. Tense About Tensors?  Think about MNIST handwritten digits Each number is 28 pixels squared There are 10 numbers, 0-9 @SWebCEO +StephenWylie #MachineLearning
    36. 36. Tense About Tensors?  Define an input tensor of shape (any batch size, 784) x = tf.placeholder(tf.float32, shape=[None, 784])  Define a target output tensor of shape (any batch size, 10) y_ = tf.placeholder(tf.float32, shape=[None, 10])  Define weights matrix (784x10) and biases vector (10-D) @SWebCEO +StephenWylie #MachineLearning
    37. 37. One-Hot: Cool To the Touch  Load the input data from tensorflow.examples.tutorials.mnist import input_data mnist = input_data.read_data_sets('MNIST_data', one_hot=True)  One-hot?!  Think about encoding categorical features: US = 0, UK = 1, India = 2, Canada = 3, …  This implies ordinal properties and confuses learners  Break encoding into Booleans:  This is where the 10-D target output tensor comes from @SWebCEO +StephenWylie #MachineLearning US = [1, 0, 0, 0] UK = [0, 1, 0, 0] Etc…
    38. 38. TensorFlow Data Structures - Placeholders  Come from inputs prior to computation  x (input picture as vector), y_ (one-hot 10-D classification vector) @SWebCEO +StephenWylie #MachineLearning [0, 0, 0, 0, 0, 0, 0, 0, … 0, 0, 0, 0, 0, 1, 1, 0, … 0, 0, 0, 0, 1, 1, 1, 0, … 0, 0, 0, 0, 1, 1, 0, 0, … 0, 0, 0, 1, 1, 1, 0, 0, … ... [0, 0, 0, 0, 1, 0, 0, 0, 0, 0] Input x y_
    39. 39. TensorFlow Data Structures – Variables  Values (i.e. model parameters) inside nodes  Used and modified by learning process  Need to be initialized with W = tf.Variable(tf.zeros([784,10])) init = tf.global_variables_initializer()  W (weights to scale inputs by), b (bias to add to scaled value) @SWebCEO +StephenWylie #MachineLearning
    40. 40. Training a Dragon, if the Dragon is a Model  Your Simple Model: y = tf.matmul(x, W) + b  Cross-entropy: distance between guess & correct answer cross_entropy = tf.reduce_mean(tf.nn. softmax_cross_entropy_with_logits(labels=y_, logits=y))  Gradient descent: minimize cross-entropy train_step = tf.train.GradientDescentOptimizer(0.5).minimize(cross_entropy)  Learning rate: 0.5 @SWebCEO +StephenWylie #MachineLearning 𝐻 𝑦′ 𝑦 = − 𝑖 𝑦𝑖 ′ log(𝑦𝑖)
    41. 41. Dragon Get Wiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiings!  Start a Session  Run global_variables_initializer  Run training for 1000 steps sess = tf.Session() sess.run(init) for i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) train_step.run(feed_dict={x: batch_xs, y_: batch_ys})  Expensive to use all training data at once!  Pick 100 random samples each step @SWebCEO +StephenWylie #MachineLearning
    42. 42. Test Flight Evaluation  Compare labels between guess y and correct y_ correct_prediction = tf.equal(tf.argmax(y,1), tf.argmax(y_,1))  Cast each Boolean result into either a 0 or 1, then average it accuracy = tf.reduce_mean( tf.cast(correct_prediction, tf.float32))  Print the final figure print(accuracy.eval(feed_dict={ x: mnist.test.images, y_: mnist.test.labels})) @SWebCEO +StephenWylie #MachineLearning
    43. 43. Demo @SWebCEO +StephenWylie #MachineLearning
    44. 44. Future Talks, Other Talks Follow me if you want to hear these!  Build a Neural Network in Python with NumPy  Build a Neural Network with nVidia CUDA Elsewhere,  Mapmaking with Google Maps API, Polymer, and Firebase  The Process Of Arcade Game ROM Hacking @SWebCEO +StephenWylie #MachineLearning
    45. 45. More Resources  Google’s “Googly Eyes” Android app [Mobile Vision API] https://github.com/googlesamples/android- vision/tree/master/visionSamples/googly-eyes  Quick, Draw! Google classification API for sketches https://quickdraw.withgoogle.com/  Making Android Apps With Intelligence, by Margaret Maynard- Reid https://realm.io/news/360andev-margaret-maynard-reid- making-android-apps-with-intelligence/ (Video + slides) @SWebCEO +StephenWylie #MachineLearning
    46. 46. Thank You @SWebCEO +StephenWylie #MachineLearning

    Notas do Editor

  • Some machine learning algorithms work better for certain types of problems than others. However, basically all of them start out with a blank slate, or perhaps random data as intermediate values, and then the dataset is analyzed in order to find which attributes of an example outcome provide the most information gain in order to make the correct classification on the training data. Learning happens iteratively over time in a manner that tries to come up with an ever more accurate answer.
  • Chat bots need machine learning in order to parse the conversation into its grammatical components, and to understand the context of the conversation so it can come up with an appropriate response.

    Pepper the robot is used for a variety of emotion-based tasks such as teaching, helping kids with autism stay focused, and training people to make standard responses to events if they are not socially adjusted.

    I hate to say this, because my mom works in Accounts Payable, but eventually her job will be automated by a system that can automatically scan an invoice, look for the amount due, and facilitate the payment over a payment network electronically. Document recognition is important for banks too, as statements which prove income for purposes of granting a loan come in many different forms.
  • HPE Haven – text analysis, speech recognition, image recognition, face detection, recommendation engine, knowledge graph analysis
    Amazon – industry-standard ML algorithms on top of your own data sets with listeners to re-evaluate upon new data (kind of like the now-deprecated Google Cloud Prediction API)
    Amazon also just rolled out SageMaker, which is a cloud-based service to help users visualize data and select the best ML algorithm, then deploy it at scale for the inference stage
    Amazon Rekognition image & video analysis that ties in with the DeepLens camera
    Azure APIs – language analytics, face & emotion & explicit content detection, speech recognition, and recommendation APIs
    IBM Watson
    Facebook – Ported Caffe for several platforms to run light & efficiently

  • Things above the line are pre-defined models that Google has already tuned for you to give what they feel are the best results. Things below the line are where you have to provide the data and possibly the tuning yourself in order to get the best outcome.

    Tensorflow Serving falls in Local because you have to set up the parameters of the infrastructure yourself, whereas Cloud ML Engine will size the model inference environment for you automatically. However, TF Serving can also be run on any hardware you have, not just inside GCP.
  • The native implementations are brought to you by the Mobile Vision API. The RESTful API which is called Cloud Vision does not feature barcode reading capabilities since that’s something a device should be able to handle on its own natively.
  • Safe search includes Adult, Medical, Spoof, and Violence.
  • No Barcode feature because why should you send images over the network just to find out if it’s a barcode when chances are it’s not.
  • Before we get the response, there’s one more detail we need to form our request: the authentication.

    Google provides several ways to allow use of the Machine Learning APIs, and the exact ways you implement them depend on what level of user data you seek to access in your application.

    The Cloud Vision API does not require you to pass along OAuth prompts to your users. You can simply make a request by finding your browser API key (usually starting with the letters AIza) and then appending that to your query string. However, this precludes you from using Google’s handy dandy API libraries for your programming language of choice. (Then again, if you’re using ALGOL 68, SmallTalk, or 6502 assembly, maybe it’s your only option.)

    To use Google’s libraries, you need to at least set up a Service Account. This type of authentication scheme is usually used for servers communicating to other servers and requires minimal end user intervention. Also, typical controls you might find on accounts in a Google Apps domain do not apply to service accounts, therefore you could be inadvertently granting your users the chance to do something dangerous like share documents and data outside of your domain if you are not careful with the permissions you grant them.

    Nevertheless, to save ourselves some time and trouble, we are going to use the basic Service Account approach for the sake of this demo. The instructions on how to create an appropriate service worker are on this workshop’s GitHub repo.
  • importanceFraction is the “fraction of importance of this salient region with respect to the original image.” Not really sure how they come to that conclusion.
  • This demo consists of running the Googly Eyes Node.js application provided in my GitHub repo. With some files in my Google Cloud Storage bucket, I could refer to that storage bucket name and the file name to load up files with human faces in them, and then this program would use the parameters returned by the Face Detection API and do some mathematics to superimpose googly eyes on top of the image using HTML5 Canvas.
  • One would presume the native libraries would be taking advantage of lower-level calls to get to the bare metal of the phone and make the analysis faster.
  • In the past, Sentiment Analysis was only available for English text. The Changelog https://cloud.google.com/natural-language/release-notes from 11/15/16 indicates that Japanese & Spanish support is available on sentiment analysis, but their How-To guide for “Analyzing Sentiment” indicates only English is supported. What’s interesting is when an English string is scored, then sent through Translate, and then you see totally different sentiment scores.
  • There are client libraries for C#, Go, Java, and other languages, plus RPC calls, available for interaction with the NL API. For RESTful calls, this is the syntax to use if you wish to just run one particular NL API query at a time.

    analyzeEntities gives you phrases in the text that are known entities, such as persons, locations, or organizations, and references the saliency of the entity to the article as well as how many times it was mentioned.
    analyzeSentiment gives you sentiment scores for the entire text body, and analyzeEntitySentiment will give you sentiment scores for text associated with each entity and its mentions. More on this in the next slide.
    analyzeSyntax gives you all the parts of speech and grammatical considerations for each word.
    classifyText will give you an array of potential categories for the text as well as the confidence of that category classification.
  • In this construct with the “annotateText” endpoint invoked, specify in the “features” object which of the three types of NLP queries you want to run.
  • The score (formerly known as polarity) is a [-1, 1] scale of the emotional sentiment (tone) of the article, from negative/scathing/blistering to positive/gushing/cheerful.

    The magnitude of the article is a scale of how emotional the document comes across. Since each expression in the text contributes to the magnitude, longer articles will have a higher magnitude, regardless of whether the tone of each individual emotional word in the text is positive or negative in sentiment.

    The small snippet of JSON here is now contained within an object called “documentSentiment”. A sibling of this object in the returned JSON structure is called “sentences”, which is an array consisting of objects containing the text and offset of each sentence, plus a “sentiment” object containing the individual magnitude and score for that piece.
  • In this demo, I simply navigated to the Web sandbox at https://cloud.google.com/natural-language/ to run various strings of text through the tool.
  • Types of models:
    - Regression: estimate a numeric answer based on the examples along a continuous curve, akin to a formula that you would solve for
    - Categorization: group possible answers into buckets that don’t have a clear distance to other buckets
  • Here again, we need to discuss authentication, because it’s a little bit stricter for the Prediction API than for Cloud Vision.

    Users must be authenticated into the Prediction API in order to use it. We can still use Service Accounts to authenticate ourselves into the app, but you have to specify a few more settings when creating them, as specified in this workshop’s GitHub repository.

    However, to push the authentication screen to your end users, go into the API Manager (at console.*developers*.google.com), open up the Credentials page, and create an OAuth Client ID credential. It will guide you through steps based on the type of devices your application is targeting, and then use the client secrets file it creates when you are defining the client object in your application. Depending on the language you are using, the objects you create to authenticate a user through a Web-based service or a standalone application might look different than that used for Service Accounts.
  • In this example, I took the data set from Problem Set #2 from my old Machine Learning homework from Fall 2010 (at http://users.eecs.northwestern.edu/~ahu340/eecs349-ps2/ ), and loaded in the train.csv file into my Google Cloud Storage Bucket. I had to massage the CSV file into the exact format it was looking for. This includes things like removing all double quotes and changing the outcome of the example to a text value rather than a numeric value so it would classify the outcome as a categorization rather than a regression, since we probably don’t want to force the values of “0” for no and “1” for yes to be the regression we solve for.

    After loading in the training data, I took some of the classification examples and pasted them into the Prediction API sample app running locally, downloadable from this workshop’s GitHub repo.

    If you want to try this in API Explorer, break out the attributes of each example into JSON in the following format (note if copying the following to watch for unintended fancy double quotes):

    {
    "input": {
    "csvInstance": [
    "104042",
    "0",
    "77",
    "m",
    "s",
    "0",
    "9",
    "7",
    "523",
    "428824",
    "24",
    "115315"
    ]
    }
    }
  • In this example, I took the data set from Problem Set #2 from my old Machine Learning homework from Fall 2010 (at http://users.eecs.northwestern.edu/~ahu340/eecs349-ps2/ ), and loaded in the train.csv file into my Google Cloud Storage Bucket. I had to massage the CSV file into the exact format it was looking for. This includes things like removing all double quotes and changing the outcome of the example to a text value rather than a numeric value so it would classify the outcome as a categorization rather than a regression, since we probably don’t want to force the values of “0” for no and “1” for yes to be the regression we solve for.

    After loading in the training data, I took some of the classification examples and pasted them into the Prediction API sample app running locally, downloadable from this workshop’s GitHub repo.

    If you want to try this in API Explorer, break out the attributes of each example into JSON in the following format (note if copying the following to watch for unintended fancy double quotes):

    {
    "input": {
    "csvInstance": [
    "104042",
    "0",
    "77",
    "m",
    "s",
    "0",
    "9",
    "7",
    "523",
    "428824",
    "24",
    "115315"
    ]
    }
    }
  • The ReLU (rectified linear unit) function is a popular function to use inside perceptrons.

    We will describe the Softmax function in great detail later.
  • The old way, initialize_all_variables(), should have been removed back in March 2017.
  • The softmax function is one that attempts to normalize the vector of guesses so that all the outputs sum to exactly 1, thus converting the output of guesses into a probability distribution. It also serves to multiplicatively increase the weight given to any particular outcome per additional unit of evidence, but also reduce that weight with each unit of evidence withheld. Thus, it is often used to represent categorical distributions. After all, we are categorizing pictures of digits as 0 through 9; that doesn’t mean they will exist linearly in space away from each other, or even in the same order, if you were to look at the values of each attribute that makes the system think it’s one particular digit compared to another.

    Cross-entropy is used as a function to compute the cost of encoding a particular event drawn from a set. The less information required to make the distinction, the better. In this case, we can also represent how far apart the guess is from the correct answer on the training set. Cross-entropy helps perform learning faster by avoiding slowdowns that could be encountered by other functions like the quadratic cost function, where learning is impaired when it gets close to the correct answer.

    Incidentally, the formula for cross-entropy has changed slightly and has been updated in this presentation. It was noted in the comments of the example code that the previous way could become numerically unstable, and so it has been changed to what is listed here. This also simplifies the model because the new function will handle taking the softmax of y.
  • The first code snippet takes the vectors and compares the first dimension of each element between y and y_. It then generates a new vector equal to the length of y (or y_ for that matter) and each result is a Boolean as to whether the things being compared were equal or not.

    The second code snippet converts each Boolean into a 0 or 1, and then takes the average of all the elements in the vector to give you accuracy as a fraction.

    The third code snippet actually causes the accuracy to be calculated. Note this used to be print(sess.run(accuracy, feed_dict=…)) and it’s still like that in the code examples, but the way listed here is in the latest documentation and it works fine.
  • In this demo, I launchTensorFlow through a Docker instance and ran my Python code that runs a previously-generated model to classify one of the images represented in one of the datasets.

    If you happened to terminate a previously-running Docker instance, you can reactivate it and get into it again with the following commands on the command line:
    ed
    docker start `docker ps -q -l`
    docker attach `docker ps -q -l`
  • Note there are several versions of Margaret’s talk available to be found on the Internet, including one from Big Android BBQ 2016 that’s a couple months newer and might have slightly different content (like how this slide deck is different from November’s deck).
    • Software

    Recomendadas

    Mais Conteúdo rRelacionado

    Semelhante a Hands-On with Google’s Machine Learning APIs, 12/3/2017

    Discover AI with Microsoft Azure
    Jürgen Ambrosi
    H2O at BelgradeR Meetup
    Jo-fai Chow
    Azure Cognitive Services
    Alexander Galkin
    Easy path to machine learning
    wesley chun
    Machine Learning on the Cloud with Apache MXNet
    delagoya
    H2O at Poznan R Meetup
    Jo-fai Chow
    Data Science Crash Course
    DataWorks Summit/Hadoop Summit
    Cognitive Toolkit - Deep Learning framework from Microsoft
    Łukasz Grala
    Kaz Sato, Evangelist, Google at MLconf ATL 2016
    MLconf
    Deep AutoViML For Tensorflow Models and MLOps Workflows
    Bill Liu
    Azure Cognitive Services: Mood Analyzer
    Sherry List
    Melbourne Cocoa Heads CoreML Presentation
    Hon Weng Chong
    And Then There Are Algorithms
    InfluxData
    Watson Visual Recognition - NYC JavaSig
    Upkar Lidder
    Track2 02. machine intelligence at google scale google, kaz sato, staff devel...
    양 한빛
    Choosing a Machine Learning technique to solve your need
    GibDevs
    H2O at Berlin R Meetup
    Jo-fai Chow
    Lucio Floretta - TensorFlow and Deep Learning without a PhD - Codemotion Mila...
    Codemotion
    H2O Deep Water - Making Deep Learning Accessible to Everyone
    Sri Ambati
    The Art of Intelligence – Introduction Machine Learning for Oracle profession...
    Lucas Jellema

    Em destaque

    What to Upload to SlideShare
    SlideShare
    Be A Great Product Leader (Amplify, Oct 2019)
    Adam Nash
    Trillion Dollar Coach Book (Bill Campbell)
    Eric Schmidt
    APIdays Paris 2019 - Innovation @ scale, APIs as Digital Factories' New Machi...
    apidays
    A few thoughts on work life-balance
    Wim Vanderbauwhede
    Is vc still a thing final
    Mark Suster
    The GaryVee Content Model
    Gary Vaynerchuk
    Mammalian Brain Chemistry Explains Everything
    Loretta Breuning, PhD
    Blockchain + AI + Crypto Economics Are We Creating a Code Tsunami?
    Dinis Guarda
    The AI Rush
    Jean-Baptiste Dumont
    AI and Machine Learning Demystified by Carol Smith at Midwest UX 2017
    Carol Smith
    10 facts about jobs in the future
    Pew Research Center's Internet & American Life Project
    Harry Surden - Artificial Intelligence and Law Overview
    Harry Surden
    Inside Google's Numbers in 2017
    Rand Fishkin
    Pinot: Realtime Distributed OLAP datastore
    Kishore Gopalakrishna
    How to Become a Thought Leader in Your Niche
    Leslie Samuel
    Visual Design with Data
    Seth Familian
    Designing Teams for Emerging Challenges
    Aaron Irizarry
    UX, ethnography and possibilities: for Libraries, Museums and Archives
    Ned Potter
    Winners and Losers - All the (Russian) President's Men
    Ian Bremmer

    Livros relacionados

    Gratuito durante 30 dias do Scribd

    Ver tudo
    Linux para principiantes Jason Cannon
    (0/5)
    Gratuito
    Arduino: Guia para colocar suas ideias em prática Fernando Bryan Frizzarin
    (4/5)
    Gratuito
    Python: Escreva seus primeiros programas Felipe Cruz
    (4/5)
    Gratuito
    O Kung Fu da Linha de Comando Jason Cannon
    (1/5)
    Gratuito
    Bitcoin Startup - Um Guia para Compreender as Finanças Descentralizadas (De.Fi), Moedas Criptográficas e Token Digitais Kevin J. Simmons
    (5/5)
    Gratuito
    SQL para Iniciantes Preston Prescott
    (3/5)
    Gratuito
    PHP: Aprenda programação PHP rápida e facilmente. Troy Dimes
    (4/5)
    Gratuito
    Técnicas de invasão de sistemas pentest: práticas detalhadas de hacker ético para os novos profissionais do mercado César Felipe Gonçalves da Silva
    (3.5/5)
    Gratuito
    A Fintech em um Flash Agustin Rubini
    (0/5)
    Gratuito
    Lógica de Programação: Crie seus primeiros programas usando Javascript e HTML Paulo Silveira
    (4/5)
    Gratuito
    NFTs Guia para Inciantes - Aprenda a entender e ganhar com i Non -Fungible Token Alessandro Hoffman
    (5/5)
    Gratuito
    A Bíblia Hacker 2021 Vol. 1 Marco Aurélio Thompson
    (3.5/5)
    Gratuito
    Metaverso - Um guia para iniciantes para a nova fronteira e como investir no novo mundo virtual em Blockchain, Criptomoeda, Arte Digital, NFT (Non-Fungible Token) e Terra Virtual Alessandro Hoffman
    (5/5)
    Gratuito
    Implementando o método GTD com o Evernote Jonathan Baker
    (4/5)
    Gratuito
    Liberdade digital: O mais completo manual para empreender na internet e ter resultados Hyeser Souza
    (5/5)
    Gratuito
    Wordpress Para Desenvolvedores Marcelo Xavier Vieira
    (5/5)
    Gratuito

    Audiolivros relacionados

    Gratuito durante 30 dias do Scribd

    Ver tudo
    Inteligência artificial: Processos de aprendizado de máquina, aprendizado profundo e automação John Adamssen
    (2.5/5)
    Gratuito
    Inteligência artificial: Análise de dados e inovação para iniciantes John Adamssen
    (4/5)
    Gratuito
    Inteligência artificial: Aprendendo sobre chatbots, robótica e outros aplicativos de negócios John Adamssen
    (4.5/5)
    Gratuito
    Divulgar site Ricardo Garay
    (0/5)
    Gratuito
    Inteligência artificial: O guia completo para iniciantes sobre o futuro da IA John Adamssen
    (5/5)
    Gratuito
    Ouvindo vozes: Como criar um podcast de sucesso e ainda ganhar dinheiro com isso Gus Lanzetta
    (0/5)
    Gratuito
    Inteligência artificial: Robôs, aplicativos e aprendizado de máquina em poucas palavras John Adamssen
    (4/5)
    Gratuito
    Tecnologia versus Humanidade: O confronto futuro entre a Máquina e o Homem Gerd Leonhard
    (4.5/5)
    Gratuito
    Treinamento cerebral: A ciência do estudo, aprendizagem e memória Dave Farrel
    (5/5)
    Gratuito
    Treinamento cerebral: Inteligência artificial e neurociência explicada Dave Farrel
    (0/5)
    Gratuito
    Treinamento cerebral: Aumente sua inteligência e memória Dave Farrel
    (0/5)
    Gratuito
    Blockchain: Aplicações no Mundo Real e Compreensão: Como a Blockchain Pode Ser Aplicada em Seu Mundo Wayne Walker
    (5/5)
    Gratuito
    Bitcoin para iniciantes e inexperientes: Criptomoedas e Blockchain Giovanni Rigters
    (5/5)
    Gratuito
    Treinamento cerebral: Aprenda mais rápido e concentre-se mais com essas técnicas Dave Farrel
    (5/5)
    Gratuito
    Inteligência artificial: Como aprendizado de máquina, robótica e automação moldaram nossa sociedade John Adamssen
    (5/5)
    Gratuito
    Inteligência artificial: Noções básicas sobre aplicativos de negócios, automação e mercado de trabalho John Adamssen
    (3/5)
    Gratuito

    ×