You are given two sections (A and B). Section A is numerical and Section B is programming. Both must be answered. For section A, you should submit handwritten answers and for section B. an annotated google colab sheet must be submitted. Only these forms of answers will be accepted. SECTION A (3 marks) An advertisement campaign is made in town. The advertiser launches 5 advertisements and will be earning dollars for each clicks. After a day, he made the following observations at different time steps. 1. Advertisement 5 earned 4 clicks 2. Advertisement 3 earned 0 clicks 3. Advertisement 2 earned 2 clicks 4. Advertisement 4 earned 6 clicks 5. Advertisement 5 earned 4 clicks 6. Advertisement 2 earned 1 clicks 7. Advertisement 1 earned 3 clicks 8. Advertisement 5 earned 7 clicks The advertisers aim is to maximize the revenue but he does not have any choice towards the choice of advertisement. 1. Figure out a strategy towards the choice of advertisement. (1 mark) 2. Analyze the results by setting initial value as 0 and initial value as 3. (1.5 mark) 3. Reason out how exploration-exploitation differs with different initial values (0.5 mark) SECTION B (4 marks) Everybody loves the treasure hunt game be it in real life or in a digital form. Suppose that there is a ForestGrid and one player who has to start his hunt from an initial position in the grid and to reach to the last grid of the ForestGrid to obtain the treasure. But unfortunately, there are certain obstacles coming his way which will lead him to a different grid rather than following an optimal path to reach near the treasure. The obstacles could be a wild animal or an end point in the forest which will get discovered later after reaching that point. For the player to reach the treasure he will have to avoid the grid containing wild animals and take another optimal move than the current one. In the scenario where the end point arrives the player will have to take 2 steps backward and find an optimal move to start his hunt.