Project Information (Fall 2020)

Group

The final project will be done in groups of (up to) 2 students. Please submit your group member information to Web Learning before {bf Dec. 2\(^{\text{nd}}\)}. Students who didn't submit their intention then will be assigned a random team.

Topic

This will be an open-topic project. You can study anything you are interested in as long as it clearly pertains to the course material. We also provide some possible topics in section 2 for your reference.

Proposal

Please submit a proposal of your final project before {bf Dec. 11\(^{\text{st}}\)}. This proposal should be no more than 1 page (not including references), describing the topic, background, related work, problem and main proposal of your project. After that, please arrange a meeting with course staff including Prof. Li, Feng Zhao and Weida between {bf Dec. 14\(^{\text{th}}\)}-{bf Dec. 16\(^{\text{th}}\)}. They can provide some advises on your project.

Presentation

We will have a poster presentation session. Please make a poster in A0-size to describe your work and submit it to Web Learning before {bf Dec. 27\(^{\text{th}}\)}. You will present your work and evaluate others in our final class on {bf Dec. 31\(^{\text{th}}\)}. The poster template will be released soon.

Final Report

Each group should submit a written report in a single {bf PDF} document with at most {bf 4 pages} before {bf Jan. 13\(^{\text{th}}\), 2021}. All related materials, e.g., codes, should also be provided.

Grades

The final project counts 20% in your final course grades. Full credits of Final project is 100 points, which include:

• Proposal: 10 points

• Discussion with class staff: 5 points

• Poster Presentation Grades (Instructor and Guests): 15 points.

• Poster Presentation Grades (Other teams): 10 points.

• Final Report Grades: 60 points.

Sample Project Ideas and Datasets

• DRIVE: Digital Retinal Images for Vessel Extraction https://drive.grand-challenge.org/DRIVE/

• Nature Language Processing: Chinese Resume Analysis. Contact TA for datasets if you are interested.

• Repeat Buyers Prediction. https://tianchi.aliyun.com/competition/entrance/231576/introduction

• CCF General Audio Classification Challenge https://www.datafountain.cn/competitions/486

• IEEE Healthcom 2020 Challenge: Depression Prediction with EEG. https://www.biendata.xyz/competition/modma/

• OpenVaccine: COVID-19 mRNA Vaccine Degradation Prediction https://www.kaggle.com/c/stanford-covid-vaccine/overview

• XPRIZE Pandemic Response Challenge: How Policies influence COVID. https://xprize.org/pandemicresponse.

Projects from Previous Years

Proposal

• The Application of Gradient-based Meta-learning Scheme on Few-shot Learning

• Career Prediction by Social Connection

• Multi-source transfer learning

• Transfer Learning from MRI Data to CT Data

Final Report

• The Application of Gradient-based Meta-learning Scheme on Few-shot Learning

• Career Analysis by Spatial-Temporal Network

• Multi-Source Learning based on a feature transferable framework

• A Gaussian Process Regression Based Approach for　Predicting Building Cooling and Heating　Consumption

• Network-based Anomaly Detection against IoT Cyber attacks using CNN-LSTM Autoencoder

Common Pitfalls

• Simply run a model from an existing work on a slightly different dataset: think of your contribution in at least one of the following areas: application, analysis, methodology and theory.

• Topic is too broad or too ambitious: reduce the project scope

• Project relies heavily on data availability/quality: use available datasets