This session will trace history of and introduce AI, machine learning
and deep learning with present day examples of success. This session
will introduce the main classes of algorithms - supervised learning,
unsupervised learning and reinforcement learning.
This session will define what is supervised learning by discussing two
use cases that led to its development. The two use cases each will
discuss regression and classification problems.
This session will contrast the requirements of supervised learning and
discuss a famous use-case that led to its development. This session will
introduce notion of clustering, the most popular unsupervised learning
algorithm.
Within the framework of supervised algorithms, a particular class of
algorithms, called deep learning, has become very successful. Deep
learning algorithms all have basically a similar looking architecture
that mimics the neurons in human brain. It is amazing that this has led
to development of many succesful architectures across many
inter-disciplinary fields.