Welcome! Here's a list of personal projects.

Comments and feedback welcome, let's learn together!

Project 1: FrozenLake (OpenAI)

A Reinforcement Learning agent I developed to solve OpenAI's challenge, description here.

Project 2: PistonBall (PettingZoo)

An attempt at a Multi-Agent Reinforcement Learning problem to coordinate 20 pistons (agents) to push the ball from right to left (of screen) ASAP.

• Referring to https://towardsdatascience.com/multi-agent-deep-reinforcement-learning-in-15-lines-of-code-using-pettingzoo-e0b963c0820b
• DONE: Cleared disk space, Troubleshoot errors, Read PistonBall environment documentation.
• NEXT STEP:
  1. Briefly read the documentation of (imports) stable_baselines3 and supersuit.
  2. Look through code to understand what is going wrong.
  3. Troubleshoot again.
• Implement, and optimize. Then apply to another problem.

Project 3: LUMO$

A Python script for automating transfer of logged expenses from HTML Telegram saved messages to Excel using BeautifulSoup.
Motivation:

1. Solves the problem of typos when I key in expenses manually.
2. Saves me 30 minutes of error-checking each month. This took 5 hours to develop, so it will be worth it in 10 months!

Project 4: inspire (v1.1)

A MERN fullstack web application to inspire users, through imagery or weblinks. Check it out here.

Motivation:

1. Sometimes we are stuck in the same loop of thinking (think: same friends, same family), and having a chance to see a fresh perspective will help to give us new ideas.