An example
on how to use the notion
of information-gain ratio when computing
the best split.
Cover Bayesian classifiers (Sect. 18.4.2.2), where
P(WZ|X)P(X)
P(X|WZ) = ------------------------
P(WZ|X)P(X) + P(WZ|~X)P(~X)
and where in the "naive" case we can assume that P(WZ|X) = P(W|X)P(Z|X).
Look at Chap 18 Slides, pp. 37, 40--41.
Embark on Chapter 15 on Transactions,
introducing the
ACID properties.
Look at Chap 15 Slides,
pp. 1--11.