Prabhat M

I am a

Quantitative Researcher | Machine Learning Engineer | Neurosymbolic AI Developer

About Me

I'm a quantitative researcher and machine learning engineer passionate about solving real-world problems through the intersection of finance, AI, and sports analytics. My work combines rigorous statistical modeling with modern deep learning to uncover structural insights from complex, noisy data.

I specialize in building hybrid AI systems that merge symbolic reasoning with neural networks, developing sophisticated financial models, and applying quantitative principles to unconventional domains like Formula 1 racing. I believe in creating models that not only predict accurately but also explain the underlying mechanisms driving performance.

Currently, I'm focused on quantitative finance research, neurosymbolic AI applications, and developing reproducible, well-documented analysis pipelines. I actively share insights and research findings through technical writing on Medium and open-source contributions on GitHub.

Areas of Expertise

Quantitative Finance

Financial modeling, algorithmic trading, risk management, performance attribution, and statistical arbitrage

Machine Learning

Deep learning, adversarial robustness, recommender systems, time series forecasting, and model interpretability

Data Engineering

Data pipelines, statistical analysis, data visualization, and reproducible research workflows

Neurosymbolic AI

Hybrid AI systems combining symbolic reasoning with neural networks for healthcare and domain-specific applications

QuantF1: Quantitative Analysis in Motorsports

A quantitative framework for understanding Formula 1 driver performance under uncertainty, published as a research series on Medium. Applying financial metrics and statistical modeling to analyze driver efficiency, risk management, and adaptability.

Six-Step Analytical Framework

Sharpe Ratio

Efficiency: How much pace does a driver extract per unit of chaos?

Sortino Ratio

Controlled Aggression: Which mistakes actually matter?

Execution Profile

Behavior: How is performance delivered?

Regime Sensitivity

Context: When does a driver's execution work?

Drawdown & Recovery

Resilience: What happens when things go wrong?

Consistency

Repeatability: Is this structural skill or situational brilliance?

Recent Research & Writing

I actively publish research and technical articles on Medium exploring quantitative finance, AI, and their applications.

Quant × Sports

The Physics of Failure: Measuring Drawdown & Recovery in Formula 1

Analyzing driver resilience through financial risk metrics applied to motorsports performance data.

Read Article
Quant × Sports

The Alpha of a Driver: Separating Skill from Machinery in Formula 1

Decomposing driver performance into skill (alpha) and machinery (beta) components using regression analysis.

Read Article
Neurosymbolic AI

Enhancing Triage Management in Healthcare with Neurosymbolic AI

Hybrid approaches combining symbolic reasoning and deep learning for improved disease prediction and medical decision-making.

Read Article
Deep Learning

Adversarial Attacks on Neural Networks

Comprehensive analysis of adversarial attack methods (FGSM, PGD, C&W) and defense strategies for robustness.

Read Article
Recommender Systems

Creating Hybrid Neural Networks for Movie Recommendations

Combining collaborative and content-based filtering using TensorFlow Recommenders for enhanced accuracy.

Read Article
Neurosymbolic AI

Revolutionizing Mocktail Creation with Neurosymbolic AI

Building personalized systems that generate novel mocktail recipes using hybrid symbolic and neural approaches.

Read Article

Featured Projects

A selection of my most impactful open-source projects and research implementations.

Core Research

QuantF1: Quantitative Finance in Motorsports

Statistical modeling of Formula 1 driver performance using financial metrics (Sharpe/Sortino ratios, alpha-beta decomposition). Published research on Medium.

Explore Repository
Quantitative Finance

Sentiment-Based Trading Strategy

News sentiment analysis using VADER and FinBERT with a custom backtesting engine. Demonstrates practical application of NLP in algorithmic trading.

Explore Repository
Neurosymbolic AI

Disease Classification & Triage

Hybrid AI system combining symbolic reasoning with deep learning for medical diagnosis and patient triage optimization.

Explore Repository
Adversarial ML

Neural Network Robustness

Comprehensive study of adversarial attacks (FGSM, PGD, C&W) and defense mechanisms for deep learning models.

Explore Repository
Recommender Systems

Movie Recommendations with Deep Learning

Hybrid neural network combining collaborative and content-based filtering using TensorFlow Recommenders framework.

Explore Repository
Machine Learning

ML Algorithms Documentation

Comprehensive repository documenting various machine learning algorithms with detailed implementations and practical applications.

Explore Repository

Let's Connect

I'm always interested in discussing research, collaborating on projects, or exploring new opportunities in quantitative finance, machine learning, and AI.