The Ancient Art of the Numerati

Abstraction: Free hands-on data mining textbook covering core ML techniques

Key points:

• Free Creative Commons (CC BY-NC 4.0) book by Ron Zacharski with Python code and datasets per chapter
• Covers recommendation systems via user-based collaborative filtering using Manhattan, Euclidean, and Minkowski distances plus Pearson correlation
• Item-based filtering chapter distinguishes explicit (star ratings) from implicit (purchase) user signals
• Classification chapters cover k-Nearest Neighbor, 10-fold cross-validation, leave-one-out, and Kappa statistic
• Naive Bayes chapter covers text classification — e.g., classifying Twitter movie reviews as positive/negative
• Final chapter covers hierarchical and k-means clustering

Connections: Ron Zacharski · Data Mining · Recommendation Systems · Classification · Clustering

Source: http://guidetodatamining.com/