Shafiq Abdul Rahman

Projects

AI Snake Game with Reinforcement Learning

Developed an AI agent to play the classic Snake game using Deep Q Learning, implemented from scratch in Python with Pygame and PyTorch.
Built and trained a neural network to predict the AI's actions, achieving consistent gameplay improvement through reinforcement learning.

Nov -2024

View Git-repo. | View training Clip

Portfolio Optimization with Graph Convolutional Networks (GCN) on Nifty 50 stocks .

Developed GCN Model: Created a GCN to capture dependencies among Nifty 50 stocks using daily returns, volatility, and RSI.
Constructed Stock Graph Applied Spectral Clustering for Diversification: Built a stock graph with correlations as edges, enhancing diversification with
Validated through Backtesting: Demonstrated effective risk‑adjusted returns, showcasing GCN’s potential in financial optimization

Nov -2024

View Git-repo.

Monte Carlo Simulation and Risk Management for Nifty 50 Portfolio Optimization

Applied Monte Carlo simulation to optimize a portfolio of Nifty 50 stocks, selecting top stocks based on Sharpe ratios.
Evaluated portfolio risk using Value at Risk (VaR) and Conditional Value at Risk (CVaR).
Identified optimal portfolio allocations to maximize returns while managing risk exposure.

May - June 2024

View Git-repo.

Interest Rate Modeling and Derivative Pricing using the Hull-White Model

Calibrated the Hull-White model parameters (mean-reversion rate and volatility) using historical interest rate data to simulate short-term interest rate paths.
Implemented Monte Carlo simulations to forecast future interest rate scenarios and used these to price interest rate derivatives (Caps and Floors)
Conducted scenario analysis by adjusting economic variables (volatility, mean reversion) to assess the impact on derivative pricing and market risks.

May 2024

View Git-repo.

Derivative instruments and their features

Investigated the trading mechanisms and valuation models of futures and options, and developed sophisticated option trading strategies.
Conducted three detailed case studies, which provided the foundation for an in‑depth, independent exploration of the Black‑Scholes and Heston model

Finlatics | Jun-Aug 2024

View Git-repo.

Optimizing Supertrend Parameters for Enhanced Profit.

Framing algorithm to optimize the indicator parameters ‑ for enhancing the profit gain using Bayesian Optimisation techniques.

Advisor : Dr.Neelesh Upadhye, IIT Madras | Jan-May 2024

View Thesis | View Slides | View Gitcode

Optimizing Supertrend Parameters for Enhanced Sharpe Ratio.

Framing algorithm to optimize the indicator parameters ‑ for enhancing the profit gain using Grid Search Optimisation techniques.

Advisor : Dr.Neelesh Upadhye, IIT Madras | Jun-Dec 2023

View Report | View Slides | View Gitcode

Linear algebra for Machine learning and Data Science

A group project, delving into linear algebra’s practical applications. Simultaneously ,explored autonomous vehicles, highlighting the dynamic link between mathematical principles and technological advancements