19. “Mobile first to AI First” Rule
User Interaction model is
fundamentally changes, all product
applies Machine Learning / Deep
learning i.e AI
20. Mobile first to AI First
People are taking the best experiences
and demanding they receive the same or
better everywhere.
- Jim Lyski, Chief Marketing Officer -
Carmax.
24. Use cases with Machine learning
❏ Recommendation - Personalize User experience.
❏ Natural Language Processing - Understanding linguistics.
❏ Financial Trading - Predicting Stock market.
❏ Marketing and Sales - Target Audience.
❏ Healthcare - Detection of Diabetic Eye Disease, Assisting Pathologists in Detecting
Cancer, Dermatology diseases classification.
Accuracy on DNA Sequencing, Drug discovery and material science.
❏ Sentiment Analysis - to understand customer’s view on product.
❏ Self-Driving Car - to avoid accident due to driver’s mistake.
❏ Object Detection - to Detect which object is there at Image / Video.
❏ Object Segmentation - to classify all the image / objects from image / videos.
❏ Dialog Agent - to ask question and get answer from specific branch of corpus.
❏ Search - to provide the information based on customer behavior.
❏ Fraud Detection - Ex. Spam Filtering
And a lot ….
25. Data Analytics Lifecycle
● Understand the Business
● Understand the Data
● Cleanse the Data
● Do Analytics the Data
● Predict the Data
● Visualize the data
● Build Insight that helps to grow Business Revenue
● Explain to Executive (CxO)
● Take Decision
● Increase Revenue
26. Machine Learning Life cycle
1. Data Quality (Removing Noisy, Missing Data)
2. Feature Engineering
3. Choosing Best Model: " based on culture of Data, For ex. If continues data-points
go with Linear Regression , If categorical binomial prediction requires then go with
Logistic Regression, For Random sample of data(Feature randomization) and have
better generalization performance. other like Gradient Boosting Trees for optimal
linear combination of trees and weighted sum of predictions of individual trees."
Try from Linear Regression to Deep Learning (RNN, CNN)
4. Ensemble Model (Regression + Random Forest + XGBoost)
5. Tune Hyper-parameters(For ex in Deep Neural Network, Needs to tune mini-batch
size, learning rate, epoch, hidden layers)
6. Model Compression - Port model to embedded / mobile devices using Compress
matrices(Sparsify, Shrink, Break, Quantize)
7. Deploy to Embedded device.
30. FUD?
Learning methods
● 30 days challenges
● Learn by doing it - Just do it
● Try understanding the math
behind it.
● Write blogs
● Join local meetup groups
● Conferences
● Read people’s code, Believe in
Open Source Contribution.
34. 1st month
Language, Practice
2nd month
Data Analysis tools
and Linear Algebra,
practice
6th month
Data visualization,
Probability
distribution, practice
10th month
Statistics, learning
techniques, Algo.
12th month
Practice, practice,
practice
35. Homework
1. No more Windows, it must be Linux.
2. Be ready with Environment - Python 2.7 and Anaconda suite. Download URL
3. Make your hands dirty with Linear Algebra and Calculus.
4. For Upcoming sessions, Attendees are advised to learn basics of Python before
attending the workshop. At the bare minimum, attendees should be knowing
Sections 1 through 5.1 in this book:
http://anandology.com/python-practice-book/