Rong Qu's BSc/MSc Project Ideas

Programmes: BSc CS, BSc CS(AI), MSc CS, MSc CS(AI), MSc DS
Project scope: Machine Learning, Data Mining and Analysis; AI Optimisation Algorithms
Example problems: transport scheduling / optimisation, resource optimisation, personnel scheduling, connected vehicles, portfolio management, credit scoring
Note: If you are interested in any of the following project ideas, please explore potential relevant datasets and prepare a draft project proposal (problem statement, datasets, techniques / algorithms) for discussion. I’m also open to other new ideas on related problem scenarios using appropriate algorithms / techniques.

• Human-in-the-loop Automated Design of Evolutionary Algorithms
  To develop an interactive tool integrating LLMs to automate the design of effective evolutionary algorithms based on existing software platforms or from scratch. Test problems include benchmark CVRPTW problem instances. Aim to deploy the tool at GitHub to support algorithm developers or researchers in transport supply chain logistics (e.g. routing). Keywords: LLM, evolutionary algorithms (local search, meta-heuristics), CVRPTW, routing algorithms Skills: programming (Python, etc.), good knowl on designing search algorithms, deployment of LLMs
• Simulation / Visualisation of Evolutionary Learning
  To develop a tool or platform visualising interactively how evolutionary algorithms solve a selected combinatorial optimisation problem (e.g. CVRPTW).
  Keywords: Evolutionary Algorithms
  Skills: programming (Python, etc.), good understanding and knowledge on evolutionary algorithms
• Multi-modal Data Analysis and Fusion of TfL Data
  To apply data analytics and fusion to the multi-modal data at the TfL datasets. Apply machine learning (e.g. clustering, random forest, ANNs) to extract knowledge, patterns or observations for potential decision makers.
  Keywords: Data analytics / mining; Data visualisation; Multi-modal Data Fusion; Machine Learning
  Skills: data analytics, programming (Jupyter Notebook, Python, Java, etc.)
• Big Data on Transport Applications
  To apply big data techniques to identify knowledge / petterns etc. in transport applications, and/or build machine learning (e.g. random forest, ANNs, regression, clustering) on large datasets, e.g. from Kaggle, UCI Machine Learning Repository, Transport of London, New York taxi, etc.
  Keywords: Artificial Intelligence; Data mining; Data visualisation
  Skills: data analytics, programming (Jupyter Notebook, Python, Java, etc.)
• Machine Learning for Driving
  To develop evolution algorithms which evolve machine learning models (e.g. neural networks, reinforcement learning) for cars to learn how to drive.
  Keywords: Evolutionary Algorithms; Machine Learning
  Skills: programming (Python or Java, etc.), Weka or R
• LLM for Automated Assessment
  To develop LLMA models fine-tuned for automated assessment / feedback for academic works.
  Keywords: LLM; Deep Learning; NLP
  Skills: Python, machine learning packages
• Your own project idea
  You are encouraged to develop your own project ideas of your own inetrests. The scope should be related to artificial intelligence or machine learning, and depends on your skills and ability to learn quickly. Individual projects are different from taught modules, as you are expected to conduct independant research.
  

Selected previous BSc/MSc projects:

You may also want to develop your own ideas based on the below previous project, provided enough NEW work is included.

• Road Accident Classification using Machine Learning
• Stock Price Prediction using Artificial Neural Network (distinction, best iTi project award)
• Calculating CAMELS Ratings using Case-Based Reasoning (distinction, best iTi project award)
• Stock Price Prediction Using Artificial Neural Networks and Support Vector Machine (distinction)
• Constraint Handling in Genetic Algorithms (distinction)
• Othello: AI Search Algorithm vs. Expert Systems (distinction)
• Genetic Algorithms to Travelling Salesman Problems (distinction)
• Traffic Flow Simulation and Diversion Modelling (distinction)
• Empirical Comparisons of Evolutionary Algorithms to the Tuning of a Chess Engine (distinction)
• Genetic Algorithms on Function Optimisation
• Hybrid AI System for Portfolio Optimisation (distinction)
• Artificial Neural Networks and Decision Tree techniques on Credit Scoring
• Interactive Course/Module Registration Assistant (group project 2008/09)

Selected project demos:

Multi-modal Journey Planner in Nottingham City Transport
An interactive route planner for passengers taking bus, tram, university Hopper bus and walk modes, optimised using intelligent hyper-heuristic algorithms. Data fusion and visualisation techniques have been utilised to analyse the multi-modal data for intelligent decision support and building the optimisation model.

AI for Car Racing (individual project 2023/24)
An engaging interactive software tool demonstrating AI techniques including Reinforcement Learning and Neuroevolution for simulated training of car racing.

Simulation and Visualisation of Taxi Trips at NYC (group project 2017/18)