Overview of Machine Learning & Feature Engineering


The Presentation inside:

Slide 0

Overview of Machine Learning & Feature Engineering Machine Learning 101 Tutorial Strata + Hadoop World, NYC, Sep 2015 Alice Zheng, Dato 1


Slide 1

About us Chris DuBois Intro to recommenders Alice Zheng Overview of ML Piotr Teterwak Intro to image search & deep learning Krishna Sridhar Deploying ML as a predictive service Danny Bickson TA Alon Palombo TA


Slide 2

Why machine learning? Model data. Make predictions. Build intelligent applications.


Slide 3

Classification Predict amongst a discrete set of classes 4


Slide 4

5 Input Output


Slide 5

Spam filtering data prediction Spam vs. Not spam


Slide 6

Text classification EDUCATION FINANCE TECHNOLOGY


Slide 7

Regression Predict real/numeric values 8


Slide 8

Stock market Input Output


Slide 9

Similarity Find things like this 10


Slide 10

Similar products Product I’m buying Output: other products I might be interested in


Slide 11

Given image, find similar images http://www.tiltomo.com/


Slide 12

Recommender systems Learn what I want before I know it 13


Slide 13

14


Slide 14

Playlist recommendations Recommendations form coherent & diverse sequence


Slide 15

Friend recommendations Users and “items” are of the same type


Slide 16

Clustering Grouping similar items 17


Slide 17

Clustering images Goldberger et al. Set of Images


Slide 18

Clustering web search results


Slide 19

Machine learning … how? Data Answers I fell in love the instant I laid my eyes on that puppy. His big eyes and playful tail, his soft furry paws, … Many systems Many tools Many teams Lots of methods/jargon


Slide 20

The machine learning pipeline I fell in love the instant I laid my eyes on that puppy. His big eyes and playful tail, his soft furry paws, … Raw data Features Models


Slide 21

Three things to know about ML Feature = numeric representation of raw data Model = mathematical “summary” of features Making something that works = choose the right model and features, given data and task


Slide 22

Feature = numeric representation of raw data


Slide 23

Representing natural text It is a puppy and it is extremely cute. What’s important? Phrases? Specific words? Ordering? Subject, object, verb? Classify: puppy or not? Raw Text


Slide 24

Representing natural text It is a puppy and it is extremely cute. Classify: puppy or not? Raw Text Sparse vector representation


Slide 25

Representing images Image source: “Recognizing and learning object categories,” Li Fei-Fei, Rob Fergus, Anthony Torralba, ICCV 2005—2009. Raw image: millions of RGB triplets, one for each pixel Raw Image


Slide 26

Representing images Raw Image Deep learning features 3.29 -15 -5.24 48.3 1.36 47.1 -1.9236.5 2.83 95.4 -19 -89 5.09 37.8 Dense vector representation


Slide 27

Feature space in machine learning Raw data ? high dimensional vectors Collection of data points ? point cloud in feature space Feature engineering = creating features of the appropriate granularity for the task


Slide 28

Crudely speaking, mathematicians fall into two categories: the algebraists, who find it easiest to reduce all problems to sets of numbers and variables, and the geometers, who understand the world through shapes. -- Masha Gessen, “Perfect Rigor”


Slide 29

Algebra vs. Geometry a b c a2 + b2 = c2 Algebra Geometry (Euclidean space)


Slide 30

Visualizing a sphere in 2D x2 + y2 = 1


Slide 31

Visualizing a sphere in 3D x2 + y2 + z2 = 1 x y z 1 1 1


Slide 32

Visualizing a sphere in 4D x2 + y2 + z2 + t2 = 1 x y z 1 1 1


Slide 33

Why are we looking at spheres? = = = = Poincare Conjecture: All physical objects without holes is “equivalent” to a sphere.


Slide 34

The power of higher dimensions A sphere in 4D can model the birth and death process of physical objects High dimensional features can model many things


Slide 35

Visualizing Feature Space


Slide 36

The challenge of high dimension geometry Feature space can have hundreds to millions of dimensions In high dimensions, our geometric imagination is limited Algebra comes to our aid


Slide 37

Visualizing bag-of-words I have a puppy and it is extremely cute


Slide 38

Visualizing bag-of-words puppy cute 1 1 1 extremely


Slide 39

Document point cloud word 1 word 2


Slide 40

Model = mathematical “summary” of features


Slide 41

What is a summary? Data ? point cloud in feature space Model = a geometric shape that best “fits” the point cloud


Slide 42

Clustering model Feature 2 Feature 1 Group data points tightly


Slide 43

Classification model Feature 2 Feature 1 Decide between two classes


Slide 44

Regression model Target Feature Fit the target values


Slide 45

Visualizing Feature Engineering


Slide 46

When does bag-of-words fail? puppy cat 2 1 1 have Task: find a surface that separates documents about dogs vs. cats Problem: the word “have” adds fluff instead of information 1


Slide 47

Improving on bag-of-words Idea: “normalize” word counts so that popular words are discounted Term frequency (tf) = Number of times a terms appears in a document Inverse document frequency of word (idf) = N = total number of documents Tf-idf count = tf x idf


Slide 48

From BOW to tf-idf puppy cat 2 1 1 have idf(puppy) = log 4 idf(cat) = log 4 idf(have) = log 1 = 0 1


Slide 49

From BOW to tf-idf puppy cat 1 have tfidf(puppy) = log 4 tfidf(cat) = log 4 tfidf(have) = 0 1 log 4 log 4 Tf-idf flattens uninformative dimensions in the BOW point cloud


Slide 50

Entry points of feature engineering Start from data and task What’s the best text representation for classification? Start from modeling method What kind of features does k-means assume? What does linear regression assume about the data?


Slide 51

Dato’s Machine Learning Platform


Slide 52

Dato’s machine learning platform Raw data Features GraphLab Create Dato Distributed Dato Predictive Services


Slide 53

Data structures for feature engineering Features SFrames SGraphs


Slide 54

Machine learning toolkits in GraphLab Create Classification/regression Clustering Recommenders Deep learning Similarity search Data matching Sentiment analysis Churn prediction Frequent pattern mining And on…


Slide 55

Demo


Slide 56

Dimensionality reduction Feature 1 Feature 2 Flatten non-useful features PCA: Find most non-flat linear subspace


Slide 57

PCA : Principal Component Analysis Center data at origin


Slide 58

PCA : Principal Component Analysis Find a line, such that the average distance of every data point to the line is minimized. This is the 1st Principal Component


Slide 59

PCA : Principal Component Analysis Find a 2nd line, - at right angles to the 1st - such that the average distance of every data point to the line is minimized. This is the 2nd Principal Component


Slide 60

PCA : Principal Component Analysis Find a 3rd line - at right angles to the previous lines - such that the average distance of every data point to the line is minimized. … There can only be as many principle components as the dimensionality of the data.


Slide 61

Demo


Slide 62

Coursera Machine Learning Specialization Learn machine learning in depth Build and deploy intelligent applications Year long certification program Joint project between University of Washington + Dato Details: https://www.coursera.org/specializations/machine-learning


Slide 63

Next up today [email protected] @RainyData, #StrataConf 11:30am - Intro to recommenders Chris DuBois 1:30pm - Intro to image search & deep learning Piotr Teterwak 3:30pm - Deploying ML as a predictive service Krishna Sridhar


×

HTML:





Ссылка: