notes from a training at work on machine learning today

"A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks T, as measured by P, improves with experience E." is a 1997 definition of machine learning by Tom M. Mitchell. In the classic example of an email spam filter being wary of emails that say "I am a prince of Nigeria." the T is classifying emails as spam or not, the P is the fraction of emails labeled correctly, and the E is watching you label emails as spam or not. Machine learning has existed for forty years in the shape of the random forest model. The linear regression of deep learning and tools like TensorFlow is a modern marvel.