Everyone can do data science with the help of tools such as: - import.io for visually scraping data from the web - Pandas to wrangle data in Python - BigML to apply machine learning to data. In this presentation , I introduce what machine learning is before moving on to a case study where I show how to build a real estate pricing model. Check out import.io's webinar for the whole thing: http://blog.import.io/post/become-a-data-scientist-in-an-hour

1. Everyone can do
data science"
import.io webinar 23/9/14
!
Louis Dorard
(@louisdorard)

4. US real estate portals:"
- Realtor
- Zillow
- Trulia
- …

5. Bedrooms Bathrooms Surface (foot²) Year built Type Price ($)
3 1 860 1950 house 565,000
3 1 1012 1951 house
2 1.5 968 1976 townhouse 447,000
4 1315 1950 house 648,000
3 2 1599 1964 house
3 2 987 1951 townhouse 790,000
1 1 530 2007 condo 122,000
4 2 1574 1964 house 835,000
4 2001 house 855,000
3 2.5 1472 2005 house
4 3.5 1714 2005 townhouse
2 2 1113 1999 condo
1 769 1999 condo 315,000

6. Bedrooms Bathrooms Surface (foot²) Year built Type Price ($)
3 1 860 1950 house 565,000
3 1 1012 1951 house
2 1.5 968 1976 townhouse 447,000
4 1315 1950 house 648,000
3 2 1599 1964 house
3 2 987 1951 townhouse 790,000
1 1 530 2007 condo 122,000
4 2 1574 1964 house 835,000
4 2001 house 855,000
3 2.5 1472 2005 house
4 3.5 1714 2005 townhouse
2 2 1113 1999 condo
1 769 1999 condo 315,000

7. Let’s create a real estate pricing
model

8. Fabien Durand
(@thefabiendurand)
www.louisdorard.com/guest/everyone-can-do-data-science-importio

10. Data Science:"
- domain knowledge
- hacking abilities
- machine learning

11. What the @#?~% is ML?

13. “Which type of email is this?
— Spam/Ham”"
-> Classification

15. “How much is this house worth?
— X $”
-> Regression

16. Bedrooms Bathrooms Surface (foot²) Year built Type Price ($)
3 1 860 1950 house 565,000
3 1 1012 1951 house
2 1.5 968 1976 townhouse 447,000
4 1315 1950 house 648,000
3 2 1599 1964 house
3 2 987 1951 townhouse 790,000
1 1 530 2007 condo 122,000
4 2 1574 1964 house 835,000
4 2001 house 855,000
3 2.5 1472 2005 house
4 3.5 1714 2005 townhouse
2 2 1113 1999 condo
1 769 1999 condo 315,000

18. ML is a set of AI techniques
where “intelligence” is built by
referring to examples

20. ??

21. “A significant constraint on
realizing value from big data
will be a shortage of talent,
particularly of people with
deep expertise in statistics
and machine learning.”
(McKinsey & Co.)

22. Making ML effortless
(Bret Victor)

24. HTML / CSS / JavaScript

25. HTML / CSS / JavaScript

26. squarespace.com

29. The two phases of machine learning:
• TRAIN a model
• PREDICT with a model

30. The two methods of prediction APIs:
• TRAIN a model
• PREDICT with a model

31. The two methods of prediction APIs:
• model = create_model(dataset)!
• predicted_output =
create_prediction(model, new_input)

32. from bigml.api import BigML
!
# create a model!
api = BigML()!
source =
api.create_source('training_data.csv')!
dataset = api.create_dataset(source)!
model = api.create_model(dataset)
!
# make a prediction!
prediction =
api.create_prediction(model, new_input)!
print "Predicted output value:
",prediction['object']['output']
http://bit.ly/bigml_wakari

34. Recap

35. • Classification and regression
• 2 phases in ML: train and predict
• Prediction APIs make it easy to build
models
• Let’s use them on real estate data to
predict price from house
characteristics

36. • Encoding domain knowledge
• Making our life easier: restricting
data to only 1 city

38. BigML!
• Look at data
• Split into training and test
• Build model from training
• Evaluate model on test
• Errors: mean absolute error (or
percentage?)

40. Other import.io + BigML use cases:!
- Predict ebook rating from description
- Predict sales of etsy stores

41. Talk at #APIconUK!
tomorrow in London

42. ML Algorithm API
Automated Pred. API
Text Classification API
Vertical Pred. API
Fixed-model Pred. API
ABSTRACTION

45. www.louisdorard.com/machine-learning-book
50% off for 24 hours with code “importio”
!
!
@louisdorard