Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Upcoming SlideShare
Machine learning in production with scikit-learn
Next

7

Share

Intro to scikit learn may 2017

Tutorial on Scikit Learn I gave at SF Data Mining meetup on May 1st 2017. Review of major parts of the Scikit-Learn API and quick coding exercise on Iris Dataset

Related Books

Free with a 30 day trial from Scribd

See all

Intro to scikit learn may 2017

  1. 1. Catalit LLC SCIKIT-LEARNTUTORIAL Francesco Mosconi SF Data Mining Meetup @ Google Launchpad May 2017 Data Weekends
  2. 2. Catalit LLC BEFORE WE START Download and install: MINICONDA PYTHON 2.7 from here: https://conda.io/miniconda.html
  3. 3. Catalit LLC INTHIS WORKSHOP • Recognize problems & choose right ML technique • Load and manipulate data with Pandas • Build classification model with Scikit-Learn • Evaluate model performance with Scikit-Learn
  4. 4. Catalit LLC MLTECHNIQUES
  5. 5. Catalit LLC
  6. 6. Catalit LLC
  7. 7. Catalit LLC
  8. 8. Catalit LLC MLTECHNIQUES CONTINUOUS CATEGORICAL SUPERVISED REGRESSION CLASSIFICATION UNSUPERVISED CLUSTERING
  9. 9. Catalit LLC TYPES OF PROBLEMS
  10. 10. Catalit LLC TYPES OF PROBLEMSSentiment Analysis Heart MonitoringBook recommendation Caption generation Human recognition
  11. 11. Catalit LLC TYPES OF PROBLEMS House price prediction Document classification Social Network Analysis
  12. 12. Catalit LLC SCIKIT-LEARN
  13. 13. Catalit LLC MODEL BUILDING 1. Collection 2. Processing 3. Model Building 4. Evaluation 5. Deployment
  14. 14. Catalit LLC BENCHMARK
  15. 15. Catalit LLC CLASSIFIERS http://www.aboutdm.com/2013/04/history-of-machine-learning.html
  16. 16. Catalit LLC
  17. 17. Catalit LLC
  18. 18. Catalit LLC New!
  19. 19. Catalit LLC
  20. 20. Catalit LLC
  21. 21. Catalit LLC
  22. 22. Catalit LLC
  23. 23. Catalit LLC PROCESSING 1. Collection 2. Processing 3. Model Building 4. Evaluation 5. Deployment
  24. 24. Catalit LLC
  25. 25. Catalit LLC
  26. 26. Catalit LLC
  27. 27. Catalit LLC Transfor mer X Transfor mer X' Estimato r X'' y
  28. 28. Catalit LLC EVALUATION 1. Collection 2. Processing 3. Model Building 4. Evaluation 5. Deployment
  29. 29. Catalit LLC
  30. 30. Catalit LLC CONFUSION MATRIX • Accuracy: Overall, how often is it correct? • (TP +TN) / total Test Negative Test Positive Condition Negative TRUE NEGATIVE FALSE POSITIVE (Type I error) Condition Positive FALSE NEGATIVE (Type II error) TRUE POSITIVE
  31. 31. Catalit LLC TRAIN -TEST SPLIT Training data Testing data Model Train Model Measure performance Alldataavailable
  32. 32. Catalit LLC
  33. 33. Catalit LLC
  34. 34. Catalit LLC ATALE OF FLOWERS https://en.wikipedia.org/wiki/Iris_flower_data_set Iris Versicolor Iris Virginica
  35. 35. Catalit LLC BINARY CLASSIFICATION Sepal Length Sepal Width Petal Length Petal Width Type Flower 1 6.2 3.4 5.4 2.3 Virginica Flower 2 5.9 3.0 5.1 1.8 Virginica Flower 3 7.0 3.2 4.7 1.4 Versicolor Features Labels Data Point
  36. 36. Catalit LLC SUPERVISED LEARNING http://www.realsafety.org/wp-content/uploads/2014/11/safety-supervisors-interaction.png
  37. 37. Catalit LLC TUTORIAL Code: dataweekends.com/ml
  38. 38. Catalit LLC THANKYOU Data Weekends Next Data Weekends Dates: 2-day Machine Learning: May 6-7 2-day Intro Deep Learning: May 20 - 21 2-day Advanced Deep Learning: Jun 3 - 4 2-day Intro Deep Learning: Jun 17 - 18
  • mailmevj

    Sep. 9, 2020
  • manojkumar4799

    Jul. 22, 2020
  • vantinhkhuc

    Apr. 27, 2020
  • ouananmohammed

    Mar. 22, 2020
  • mansworld1004

    Nov. 19, 2019
  • Lumipanda

    Aug. 17, 2017
  • mikepham12

    Jul. 6, 2017

Tutorial on Scikit Learn I gave at SF Data Mining meetup on May 1st 2017. Review of major parts of the Scikit-Learn API and quick coding exercise on Iris Dataset

Views

Total views

814

On Slideshare

0

From embeds

0

Number of embeds

39

Actions

Downloads

7

Shares

0

Comments

0

Likes

7

×