Projects & Experience

• Designed and implemented DL/ML models to predict price movements of financial symbols.

• Developed an agentic framework for ticker-based sentiment analysis, enabling actionable insights.

• Designed and implemented rule-based and LLM-based trading strategies with explainability.

• Designed a backtesting framework for evaluating trading strategies.

• Implemented various algorithms from the literature for forecasts post-processing.

• Improved data pipelines efficiencies and speed.

• Worked on midterm electric load forecasting by implementing and validating a hybrid ETS+RD-LSTM model.

• Conducted an in-depth literature review of state-of-the-art forecasting methods, identifying key limitations and opportunities for improvement.

• Enhanced the baseline model’s architecture and performance through iterative experimentation and optimization.

• Led a comprehensive ablation study to understand the individual contributions of each model component, and benchmarked performance against other advanced deep learning and statistical models.

Authors: Barkous, H., Amayri, M., & Bouguila, N.

Developed a solution for extracting structured data from PDF documents.

Predicted early learning outcomes based on provided datasets.

Forecasted carbon dioxide levels using time series data.