ML 3
ML 3
4.a) What are recurrent networks? How are they get trained?
dse
b) Describe the approaches for comparing two machine learning algorithms when limited
data is available. [8+7]
Pda
5.a) State Bayes theorem and explain Naïve Bayes classifier.
b) Discuss the relationship between the maximum likelihood hypothesis and the least-
squared error hypothesis in Bayesian learning. [7+8]
pPea
6.a) What is PAC learnability? How to verify it?
b) “The distance-weighted k-NN algorithm is a highly effective inductive inference
method for many practical problems.” Support this statement with suitable discussions.
rpse
[8+7]
7.a) What is the motivation for genetic algorithms? Illustrate offspring generation from
2rs0
given population.
b) Describe the sequential covering algorithm for learning a disjunctive set of rules. [7+8]
8.a) Differentiate between inductive learning and analytical learning. How to combine these
22
two learning approaches.
b) Give the remarks on explanation based learning. [7+8]
---oo0oo---