Term papers
Each student will write a term paper on one of the topics listed below. The term paper should cover all the key aspects of the topic and contain all the important references related to it.
Suggested Topics
- The importance of the Gibbs sampler in Machine-Learning.
- Applications of random walks in data mining and information retrieval.
- Efficient computation of Pagerank.
- The use of random walks for graph clustering.
- Applications of Pagerank beyond Web search.
- Locality-sensitive hashing and its applications.
- Applications of SVD in Machine Learning.
- Applications of clustering in Machine Learning.
General guidelines
Imagine a 2nd year computer science undergraduate student is your option. This person has a basic mathematical training, knows programming and has studied data structures, but has yet to study any advanced material. Can you make the topic accessible to this individual? How useful is your term paper for this person trying to learn about this topic? Here are some suggestions to help you along
- Read widely about the topic from textbooks and research papers. Make a list of all sources you have read and keep links to them available.
- Organize your term paper as follows:
- Introduction. Give an overview of what you are going to discuss in the term paper. Write this section last.
- Basic definitions. Introduce the basic concepts required to understand what you're talking about, e.g., define SVD or pagerank or whatever, and make other definitions required to understand your definition. Assume basic knowledge.
- Other sections. Organize what you want to say in terms of sections. Here you may talk about different categories of applications/algorithms/methods in different subsections. You can have as many of these middle sections as possible.
- Conclusions and Future directions. Here you must talk about what are the open issues in your topic and what scope for further research you see.
- Use latex for your paper (mandatory). Use bibtex for your references.
- Make sure each citation is complete and includes author names, full paper title, where published (conference/journal/arXiv/internet), page numbers if known, year. Missing any of these except page numbers is not acceptable.
- Finally, once written, re-read and revise. Try to imagine how a person who is reading this for the first time will feel and revise to make it more comprehensible for this person.
Evaluation
- First draft due on 16 April 2021. (30% weightage)
- Final draft due on 7 May 2021. (70% weightage)
All submissions will be through Moodle.
Amitabha
Bagchi