Technical Expertise

Predictive Modelling

Deep Learning & Large-Scale Computing

Machine Learning & Classification

Logistic Regression: Used for classification tasks, assessing binary outcomes with high accuracy, commonly for risk assessment and customer chum prediction.
GARCH : Applied for volatility forecasting in financial markets, with further applications in economic data series to quantify risk via Value-at-Risk (VaR).
Poisson & Zero-Inflated Poisson Models: Modelled discrete outcomes, particularly useful in event-count forecasting for scenarios with over-dispersed or zero-inflated data.

Machine Learning & Classification

Deep Learning & Large-Scale Computing

Machine Learning & Classification

K-Nearest Neighbors (KNN): Applied to structured datasets for customer segmentation and personalized recommendations, optimized through hyperparameter tuning and cross-validation.
Support Vector Machines (SVM): Implemented for boundary-based classifications, delivering robust performance in high-dimensional spaces, often used in credit risk and fraud detection.
Random Forests & Decision Trees: Used for both classification and regression, achieving high predictive accuracy through ensemble learning techniques, suitable for high-dimensional financial and customer data.

Deep Learning & Large-Scale Computing

Deep Learning & Large-Scale Computing

Bayesian Inference & Probabilistic Modelling

TensorFlow Neural Networks: Developed predictive health and financial models using neural network architectures, enhanced with optimization algorithms like Gradient Descent and Adam for efficient learning.
CNNs (Convolutional Neural Networks): Utilized for advanced image and pattern recognition across large datasets, enabling analysis for visual data and anomaly detection in time-series.
LSTM (Long Short-Term Memory Networks): Applied for sequential data forecasting, particularly useful for time-series data and trend analysis in financial and climate models.

Bayesian Inference & Probabilistic Modelling

Bayesian Inference & Probabilistic Modelling

Bayesian Inference & Probabilistic Modelling

MCMC (Markov Chain Monte Carlo): Applied to estimate parameters in uncertain environments, capturing changes in patterns within time-series data, as seen in market trend and climate change analyses.
Metropolis-Hastings Algorithm: Used in Bayesian computation to sample from complex, high-dimensional distributions, providing credible intervals and probabilistic forecasts.

Financial Modelling & Portfolio Optimization

Bayesian Inference & Probabilistic Modelling

Financial Modelling & Portfolio Optimization

Fama-French Three-Factor Model: Developed asset pricing models for portfolio construction, analyzing risk premiums and ensuring balanced asset allocation.
Monte Carlo Simulations: Used for risk and return scenario analyses, simulating asset behaviour under various market conditions to support decision-making for investments and capital allocation.
Black-Scholes & Option Pricing Models: Implemented for derivative pricing, calculating fair values under different market conditions to manage financial instruments and hedge risks..

Portfolio

Auto Insurance Fraud Prediction

Auto Insurance Fraud Prediction

Auto Insurance Fraud Prediction

DeMark Dashboard: TD Sequential

Auto Insurance Fraud Prediction

Auto Insurance Fraud Prediction

Shiny: UK Climate Dashboard

Auto Insurance Fraud Prediction

Anticipating Volatility: Commodities

Anticipating Volatility: Commodities

Time Series Forecasting: Air Pollution

Anticipating Volatility: Commodities

Time Series Forecasting: Air Pollution

Time Series Forecasting: Air Pollution

Time Series Forecasting: Air Pollution

Bayesian Changepoint Detection

Time Series Forecasting: Air Pollution

Time Series Forecasting: Air Pollution