Practical Machine Learning and Rails
summary
Shows concrete pipelines: sentiment classifier built from tweets labeled by emoticons, converted to bag-of-words (1,000-word dictionary) and cross-validated to about 76% accuracy using Weka + liblinear with JRuby integration. Covers feature work: bigrams/triagrams, TF/IDF, stemming, attribute selection and avoiding overfitting. Gives a regression example for domain-price prediction using per-word features (Google search volume, AdWords CPC) fed to support-vector regression. Tools, dataset sizes, and measured results are all explicit.
AI-written summary · about
chapters (5)
- 0:00Introduction to machine learning in Rails
- 15:56Sentiment analysis emoticons
- 27:16Weka sentiment classifier
- 35:21Feature engineering for predictions
- 43:40Q&A
related talks
35mIntelligent Ruby: Getting Started with Machine Learning
48mLearning Machine Learning
29mJRuby on Elephants
27mFive machine learning techniques that you can use in your Ruby apps today By Benjamin Curtis Machine learning is everywhere these days
28mMachine Learning for Fun and Profit
51mMachine Learning with Ruby
37mLearning Machine Learning
30mMachine Learning: Beyond the Tutorial