ACADEMIC PROJECTS Link to heading

NTU EEE Final Year Project (Aug 2023 – May 2024) Link to heading

Intelligent Robot Manipulation with Deep Reinforcement Learning

  • Engaged in a collaborative research initiative with A*STAR SIMTech.
  • Focused on applying reinforcement learning algorithms for control systems in robotics, particularly a 6-DOF (Degrees of Freedom) robotic arm.
  • Developed a novel teacher-student framework to mitigate the simulation-to-reality gap in robotic learning, leveraging both reinforcement learning and imitation learning for effective knowledge transfer from simulation (teacher) to real-world scenarios (student).

NTU-EEE Design and Innovation Project (Aug 2022 – Nov 2022) Link to heading

APP Development of Scientific Calculator (Team of 8)

  • Gained proficiency in Flutter and Dart to develop a calculator mobile application.
  • Integrated a JavaScript symbolic math library for advanced calculations, including calculus, complex number operations, and symbolic math manipulations.
  • Awarded first place in the category.

PERSONAL PROJECTS Link to heading

Cryptocurrency Algorithmic Trading (August 2023 – Present) Link to heading

  • Independently developed and tested cryptocurrency trading strategies using Python on open-source framework (FreqTrade), incorporating machine learning and reinforcement learning techniques.
  • Implemented strategies based on the latest academic research.
  • Deployed backtested strategies on cloud computing platforms for testing and optimization.

Deep Reinforcement Learning Project with Unity-Based Environment (21 – 31 October 2023) Link to heading

  • Trained a virtual agent to navigate and find its target in a simulated 3D environment using the Unity engine.
  • Utilized the Proximal Policy Optimization (PPO) algorithm for training.
  • Optimized the training process to complete within 1 hour through hyperparameter tuning.

COMPETITION Link to heading

NTU UAVionics Club Robotics Car Competition (Team of 3) (1 – 4 March 2023) Link to heading

  • Designed and constructed a controllable robotics car using components such as ESP32-Devkit, IMU, LiDAR sensor, and encoders.
  • Developed and programmed the robotics car for autonomous navigation using Arduino and ROS.
  • Secured first place in the competition.

NTU TradeMaster Cup 2022 (Nov 2022 - Jan 2023) Link to heading

  • Engaged in the TradeMaster Cup 2022 to develop profitable reinforcement learning agents for portfolio management in stochastic financial markets.
  • Tasked with optimizing a reinforcement learning (RL) agent’s accumulative return using financial market data, demonstrating the application of RL in quantitative finance.
  • Gained insights into reinforcement learning’s potential to solve complex financial decision-making problems, despite not securing a winning position.