13.07.2015 Views

Data Mining: Practical Machine Learning Tools and ... - LIDeCC

Data Mining: Practical Machine Learning Tools and ... - LIDeCC

Data Mining: Practical Machine Learning Tools and ... - LIDeCC

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

xCONTENTS5 Credibility: Evaluating what’s been learned 1435.1 Training <strong>and</strong> testing 1445.2 Predicting performance 1465.3 Cross-validation 1495.4 Other estimates 151Leave-one-out 151The bootstrap 1525.5 Comparing data mining methods 1535.6 Predicting probabilities 157Quadratic loss function 158Informational loss function 159Discussion 1605.7 Counting the cost 161Cost-sensitive classification 164Cost-sensitive learning 165Lift charts 166ROC curves 168Recall–precision curves 171Discussion 172Cost curves 1735.8 Evaluating numeric prediction 1765.9 The minimum description length principle 1795.10 Applying the MDL principle to clustering 1835.11 Further reading 1846 Implementations: Real machine learning schemes 1876.1 Decision trees 189Numeric attributes 189Missing values 191Pruning 192Estimating error rates 193Complexity of decision tree induction 196From trees to rules 198C4.5: Choices <strong>and</strong> options 198Discussion 1996.2 Classification rules 200Criteria for choosing tests 200Missing values, numeric attributes 201

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!