COMP90051: Statistical Machine Learning - Implement ML Approaches in Code - IT Assignment Help

Assignment Task:

Multi-armed bandits (MABs) are a simple but powerful framework for sequential decision making under uncertainty. Through the 2000s, Yahoo! Research led the way in applying MABs to problems in online advertising, information retrieval, and media recommendation. One of their many applications was to Yahoo! News, in deciding what news items to recommend to users based on article content, user profile, and the historical engagement of the user with articles. Given decision making in this setting is sequential (what do we show next?) and feedback is only available for articles shown, Yahoo! researchers observed a perfect formulation for MABs like those (-Greedy and UCB) learned about in class. Going further, however, they realised that incorporating some element of user-article state requires contextual bandits: articles are arms; context per round incorporates information about both user and article (arm); and {0, 1}-valued rewards represent clicks. Therefore the per round cumulative reward represents click-through-rate, which is exactly what services like Yahoo! News want to maximise to drive user engagement and advertising revenue. In this project, you will work individually (not in teams) to implement several MAB algorithms. Some will be directly from class, while others will be more advanced and come out of papers that you will have to read and understand yourself.

By the end of the project you should have developed:

ILO1. A deeper understanding of the MAB setting and common MAB approaches;

ILO2. An appreciation of how MABs are applied;

ILO3. Demonstrable ability to implement ML approaches in code; and

ILO4. An ability to pick up recent machine learning publications in the literature, understand their focus, contributions, and algorithms enough to be able to implement and apply them. (And being able to ignore other presented details not needed for your task.)

Overview

You will be completing the following tasks, noting that you need not complete the entire project to achieve a high mark:

1. Implement -greedy and UCB MABs (StatML lecture 16) [4 marks]

2. Implement off-policy evaluation (Li et al., 2010, 2011) [4 marks]

3. Implement LinUCB contextual MAB (Li et al., 2010) [7 marks]

4. Implement TreeBootstrap contextual MAB (Elmachtoub et al., 2017) [7 marks]

5. Evaluation and hyperparameter tuning for LinUCB [3 marks]

6. Implement KernelUCB contextual MAB (Valko et al., 2013) [5 marks]

All tasks are to be completed in the provided Python Jupyter notebook proj2.ipynb. 2 Detailed instructions for each task are included at the end of this document. Most tasks will require you to consult one or more references (e.g. academic papers or lecture slides)—we provide helpful pointers to guide your reading and to correct any ambiguities.

This COMP90051: IT Assignment has been solved by our IT Experts at onlineassignmentbank. Our Assignment Writing Experts are efficient to provide a fresh solution to this question. We are serving more than 10000+ Students in Australia, UK & US by helping them to score HD in their academics. Our Experts are well trained to follow all marking rubrics & referencing style.

Be it a used or new solution, the quality of the work submitted by our assignment experts remains unhampered. You may continue to expect the same or even better quality with the used and new assignment solution files respectively. There’s one thing to be noticed that you could choose one between the two and acquire an HD either way. You could choose a new assignment solution file to get yourself an exclusive, plagiarism (with free Turnitin file), expert quality assignment or order an old solution file that was considered worthy of the highest distinction.