Summary

TABLE OF CONTENTS

• Getting Started

  • Installation
  • Install as permanent functions
  • Input Format

• List of Functions

• Tips for Effective Hivemall

  • Explicit add_bias() for better prediction
  • Use rand_amplify() to better prediction results
  • Real-time prediction on RDBMS
  • Ensemble learning for stable prediction
  • Mixing models for a better prediction convergence (MIX server)
  • Run Hivemall on Amazon Elastic MapReduce

• General Hive/Hadoop Tips

  • Adding rowid for each row
  • Hadoop tuning for Hivemall

• Troubleshooting

  • OutOfMemoryError in training
  • SemanticException generate map join task error: Cannot serialize object
  • Asterisk argument for UDTF does not work
  • The number of mappers is less than input splits in Hadoop 2.x
  • Map-side join causes ClassCastException on Tez

Part II - Generic Features

• List of Generic Hivemall Functions
• Efficient Top-K Query Processing
• Text Tokenizer
• Approximate Aggregate Functions

Part III - Feature Engineering

• Feature Scaling
• Feature Hashing
• Feature Selection
• Feature Binning
• Feature Paring
  • Polynomial features
• Feature Transformation
  • Feature vectorization
  • Quantify non-number features
  • Binarize label
  • One-hot encoding
• Term Vector Model
  • TF-IDF Term Weighting
  • Okapi BM25 Term Weighting

Part IV - Evaluation

• Binary Classification Metrics
  • Area under the ROC curve
• Multi-label Classification Metrics
• Regression Metrics
• Ranking Measures
• Data Generation
  • Logistic Regression data generation

Part V - Supervised Learning

• How Prediction Works
• Step-by-Step Tutorial on Supervised Learning

Part VI - Binary Classification

• Binary Classification

• a9a Tutorial

  • Data Preparation
  • General Binary Classifier
  • Logistic Regression
  • Mini-batch Gradient Descent

• News20 Tutorial

  • Data Preparation
  • Perceptron, Passive Aggressive
  • CW, AROW, SCW
  • General Binary Classifier
  • AdaGradRDA, AdaGrad, AdaDelta
  • Random Forest
  • XGBoost

• KDD2010a Tutorial

  • Data Preparation
  • PA, CW, AROW, SCW

• KDD2010b Tutorial

  • Data Preparation
  • AROW

• Webspam Tutorial

  • Data Pareparation
  • PA1, AROW, SCW

• Kaggle Titanic Tutorial

• Criteo Tutorial

  • Data Preparation
  • Field-Aware Factorization Machines

Part VII - Multiclass Classification

• News20 Multiclass Tutorial

  • Data Preparation
  • Data Preparation for one-vs-the-rest classifiers
  • PA
  • CW, AROW, SCW
  • XGBoost
  • Ensemble learning
  • one-vs-the-rest Classifier

• Iris Tutorial

  • Data preparation
  • SCW
  • Random Forest
  • XGBoost

Part VIII - Regression

• Regression

• E2006-tfidf Regression Tutorial

  • Data Preparation
  • General Regessor
  • Passive Aggressive, AROW
  • XGBoost

• KDDCup 2012 Track 2 CTR Prediction Tutorial

  • Data Preparation
  • Logistic Regression, Passive Aggressive
  • Logistic Regression with amplifier
  • AdaGrad, AdaDelta

Part IX - Recommendation

• Collaborative Filtering

  • Item-based Collaborative Filtering

• News20 Related Article Recommendation Tutorial

  • Data Preparation
  • LSH/MinHash and Jaccard Similarity
  • LSH/MinHash and Brute-force Search
  • kNN search using b-Bits MinHash

• MovieLens Movie Recommendation Tutorial

  • Data Preparation
  • Item-based Collaborative Filtering
  • Matrix Factorization
  • Factorization Machine
  • SLIM for fast top-k Recommendation
  • 10-fold Cross Validation (Matrix Factorization)

Part X - Anomaly Detection

• Outlier Detection using Local Outlier Factor (LOF)
• Change-Point Detection using Singular Spectrum Transformation (SST)
• ChangeFinder: Detecting Outlier and Change-Point Simultaneously

Part XI - Clustering

• Latent Dirichlet Allocation
• Probabilistic Latent Semantic Analysis

Part XII - GeoSpatial Functions

• Lat/Lon functions

Part XIII - Hivemall on SparkSQL

• Getting Started

  • Installation

• Binary Classification

  • a9a Tutorial for SQL

• Regression

  • E2006-tfidf Regression Tutorial for SQL

Part XIV - Hivemall on Docker

• Getting Started

Part XIV - External References

• Hivemall on Apache Pig