Machine Learning with
  _____       _           
 |  __ \     | |          
 | |__) |   _| |__  _   _ 
 |  _  / | | | '_ \| | | |
 | | \ \ |_| | |_) | |_| |
 |_|  \_\__,_|_.__/ \__, |
                     __/ |
                    |___/     Programming Language

Saleem Ansari (@tuxdna)

http://tuxdna.in/

What is Machine Learning?

Some Math

Demo

• 1959 - Field of study that gives computers the ability to learn without being explicitly programmed.

• 1998 - A computer program is said to learn from experience E with respect to some task T and some performance P, if its performance on T, as measured by P, improves with experience E.

• Product Recommendation: understanding / inferring what your customers are looking for
• Clustering: grouping similar items or grouping very similar documents, which are perhaps talking about the same subject
• Regression and Classification: predicting house prices, or identifying a class of an item viz. product, document, person etc.
• Topic Modeling: identifying topics from documents
• Frequent Patterns Mining: knowing which entities occur together very often
• And many more

• Similarity and Distance metrics
• Vector and Matrices
• Statistics
• Probability

Different Similarity metrics

• Pearson correlation
• Euclidean distance
• Cosine measure
• Spearman correlation
• Tanimoto coefficient
• Log likelihood test

Distance to Similarity conversion ( not the only way )

s = 1 / ( 1 + d )

What are the stats almost everyone knows?

• mean / average / expectation
• median
• mode

What about these?

• variance
• stardard deviation

• Conditional Probability: P(A|B) = num(A intersection B) / num(B)
• Bayes Rule: P(A|B) = P(B|A) / P(B)
• Probability Distribution: PMF for discreet, PDF for continuous variables

• Vector
• Matrix

( see the ruby docs )

• What about a SparseMatrix ?

• Generating Recommendations on MovieLens data
• Statistical News Classifier
• Clustering the Reuters news data
• Naive Bayes Classifier

Challenges to be solved within Ruby ecosystem:

• Fast Math
• Easy Plotting
• Integrated Environment for ML

Is there any hope ?

• AI4R
• SciRuby
• NMatrix
• JRuby and Apache Mahout

• Machine Learning Intro
• ML Intro Code
• News Aggregator App
• Intro to ML in Ruby
• Thinking about ML in Ruby
• ML is way easier that it looks
• Text Classification using SVM
• ML with Ruby
• SVM in Ruby
• Weka ML Toolkit
• Weka Explorer
• Data Mining with Weka

• SVM
• Selecting an appropriate similarity metric & assessing the validity of a k-means clustering model
• Converting Similarity to Distance and vice-versa
• Mahout Scala and Spark bindings
• Multivariate Normal Distribution
• Recommendation Engine
• Logistic Regression
• Practical Machine Learning