 Hello, my name is Professor Manjesh Kumar Hanwal. I am from the Department of Industrial Engineering Operations Research. I am going to take this course named online machine learning. Okay, about this course, often in life we have to make decisions in an uncertain environment and the outcome of our decisions in such uncertain environments could also be uncertain. But still in such cases, we would like to identify a decision or a available action to us that performs well on an average. For example, let us say you have a set of drugs available to cure a disease. Apparently, we do not know which is the most effective drug among all of this, but you would like to experiment and quickly identify the drug which gives this disease well on an average. Other example is, suppose you are an advertisement, you are in the advertisement industry and you want to, you have a slot in which you want to put an advertisement from a available set of advertisements. Here, your goal would be like which advertisement to put in this slot so that it will be clicked by most number of users. So this course is all about decision making in uncertain environments. In this course, we will start with study of adversarial settings where the rewards assigned by the environment could be arbitrary. It need not follow any particular distributions. Then we focus on the stochastic case where the rewards or let us say losses are assigned according to a distribution which is unknown to us but fixed. Then we move on to the study of contextual cases, contextual bandit cases where we would like to identify an action which is optimal for each possible context we are going to see. At the end, we will study a set of problems called pure exploration problems where our goal would be to identify the best action after doing certain number of explorations. But here, we would like to keep exploration as low as possible but still we would like to identify the best action. So overall, this course is about study of various learning environments, come up with appropriate algorithms and give their performance guarantees. Thank you.