Page 138 - Artificial Intellegence_v2.0_Class_12
P. 138
Unsupervised Learning Algorithm
K Means Clustering
efore eans After eans
eans
his unsupervised learning algorithm divides data sets into a predetermined number of clusters in a ay that makes
each cluster's data homogeneous and distinct from the data in the other clusters.
Other examples include – Anomaly detection, Neural Networks
Reinforcement Learning Algorithm
Q-Learning
nvironment
Action tate
Agent e ard
his is a model free, off policy algorithm and is based on the ellman quation. he technique makes use of
a table, a lookup database that records the agent's estimated utility or quality of doing a specific action in a
specific condition. hrough trial and error, the agent updates the values to ma imise it in the table and eventually
converges to the best possible policy.
Other examples include – SARSA (State–action–reward–state–action), DQN (Deep Q Network)
Showcase through a Compelling Story
he follo ing story illustrates through a series of visuals ho a pro ect made a difference in the lives of thousands
of Indian omen.
136 Touchpad Artificial Intelligence (Ver. 2.0)-XII

