Scikit-Learn to "learn them all" @ EuroPython 2014

**Scikit-learn** is a powerful Python library, providing implementations of
many of the most popular machine learning algorithms.
This talk will provide a general overview of the main "batteries" included in
Scikit-learn, along with working code examples and comparisons with other
existing machine learning tools, in order to get the best from this wonderful
package for our machine learning code.

Transcript

1. SCIKIT-LEARN “to learn them all” MACHINE L E A R N I N G Speaker Valerio Maggio valeriomaggio@gmail.com @leriomaggio +ValerioMaggio
2. SCIKIT-LEARN TO “LEARN THEM ALL” WHY SCIKIT-LEARN is so cool in other words
3. PLEASE ANSWER TO FIVE QUESTIONS THREE QUESTIONS • Do you already know what Machine Learning is? • Have you ever used 
 Scikit-Learn? • How many of you also attend the Scikit-Learn training?
4. WHAT DOES MACHINE LEARNING MEANS?
5. WHAT IS MACHINE LEARNING ? Machine learning teaches machines how to carry out tasks by themselves. It is that simple. The complexity comes with the details. W. Richert & L.P. Coelho, 2013 Building Machine Learning Systems with Python
6. ML AT A GLANCE
7. ML & DATA ANALYSIS • Data Analysis • Data Mining • Big Data • Data Science BUZZ W O R D S
8. DATA SCIENCE DATA S C I E N C E Data science is the study of the generalizable extraction of knowledge from data Drew Conway 
 Data Science 
 Venn Diagram
9. MACHINE LEARNING & DATA ANALYSIS ML IS ABOUT 
 MAKING PREDICTIONS
10. THE ESSENCE OF MACHINE LEARNING • A Pattern exists • We cannot pin it down mathematically • We have data on it. Learning by Examples
11. MACHINE LEARNING @jakevdp
12. EXAMPLE(1): CLASSIFICATION SUPERVISED LEARNING
13. EXAMPLE(2): CLUSTERING UNSUPERVISED LEARNING
14. MACHINE LEARNING @jakevdp
15. EXAMPLE(3): CLUSTERING UNSUPERVISED LEARNING
16. WHAT ABOUT PYTHON?
17. PYTHON and DATA SCIENCE DATA SCIENCE P Y T H O N experfy.com/blog/ python-data-science/
18. PYTHON & S C I E N C E
19. ONE PYTHON to rule them all • Python: The language of choice for Data Science • Displacing R & Matlab http://goo.gl/17XA5J One of the biggest benefits of doing data science in Python is added efficiency of using one programming language across different applications. T. Yarkoni, Univ. Texas DATA SCIENCE P Y T H O N
20. DATA SCIENCE P Y T H O N
21. MACHINE LEARNING IN PYTHON
22. ML PYTHON POWERED MACHINE LEARNING P Y T H O N github.com/kevincobain2000/awesome-machine-learning
23. MACHINE LEARNING P Y T H O N Scala early stage + C++ PyWrap + Python Powered
24. Pure Python Numpy + Scipy
25. alpha
26. NLP & Machine Learning Supervised Learning (SVM) Neural Networks General Purpose
27. (General purpose) MACHINE LEARNING LIBRARIESwritten in Python
28. SO WHY TO CHOOSE
 SCIKIT-LEARN?
29. WHY SCIKIT-LEARN
30. WHY SCIKIT-LEARN
31. pip install numpy pip install scipy pip install ipython pip install scikit-learn + + + + pip install matplotlib INSTALLATION https://store.continuum.io/cshop/anaconda/
32. SCIKIT L E A R N DESIGN PHILOSOPHY • Includes all the batteries necessary for (general purpose) 
 Machine Learning Code • Data (and Datasets) • Feature Selection, Extraction algorithms • ML Algorithms (Classification, Regression, Clustering, ….) • Evaluation functions (Cross Validation, Confusion Matrix) Algorithm Selection Philosophy: Try to keep the core as light as possible, including well-known and largely used ML methods
33. SCIKIT AT A GLANCE
34. SCIKIT A P I
39. SCIKIT in action
40. DATA REPRESENTATION MACHI NE L E A R N I N G N = Number of Samples in the Data Set D = Number of Features In SCIKIT: scipy.sparse matrices
41. IRIS DATASET Goal:
 Design an algorithm that is able to automatically recognize IRIS species
42. IRIS DATASET Goal:
 Design an algorithm that is able to automatically recognize IRIS species
44. SUPERVISED LEARNING KNN CLASSIFIER
45. EXAMPLE: K=6 (6NN) Virginica Setosa Versicolor = ?
46. KNN IN SCIKIT
47. UNSUPERVISED LEARNING
48. K-MEANS IN SCIKIT Ground Truth K-Means (k=3)
49. MODEL EVALUATION How do we know if our prediction model is good? SCIKIT B A T T E R I E S
50. MODEL VALIDATION
51. CROSS VALIDATION Cross-validation is a model validation technique for assessing how the results of a statistical analysis will generalize to an independent data set.
52. SCIKIT B A T T E R I E S LARGE SCALE out-of-the-box Clustering KMeans(n_clusters=3, copy_x=True, n_jobs=-1)
53. SCIKIT meets
54. SCIKIT meets nltk.org/api/nltk.classify.html#module-nltk.classify.scikitlearn nltk.classify.scikitlearn
55. IN CONCLUSIONS MACHINE LEARNING P Y T H O N • Scikit-learn is not the only Machine Learning library available in Python • But it is powerful, and easy-to-use. • Very efficient implementations • numpy, scipy, cython • Highly Integrated • NLTK, Scikit-Image
56. LOOKING FORWARD TO USING IT, RIGHT?
57. THANKS FOR YOUR KIND ATTENTION valeriomaggio@gmail.com @leriomaggio +ValerioMaggio