>If you could name three of them I'd be really grateful. Latent Dirichlet Allocation, message-passing in Hidden Markov models, and Naive Bayes for spam filteri

>If you could name three of them I'd be really grateful. Latent Dirichlet Allocation, message-passing in Hidden Markov models, and Naive Bayes for spam filtering. Outside my subfield, there's always the basic handwriting recognition employed in ATMs.