Salem Nkunda Nyisingize

Welcome

Mathematics & computer science applied to data

Université Laval — Québec

Math + code to understand, forecast, and secure real-world systems.

New: GPS trajectory — cinematic visualization: time-based interpolation and trajectory staging.

New: RNNs, GRUs and LSTMs — understanding memory and recurrence for time series forecasting.

GPU-accelerated visualization of Mandelbrot and Julia sets. Infinite zoom into complex dynamics z² + c.

Streamlit application to explore and forecast time series (ARIMA/SARIMA, multivariate methods, deep learning).

Interactive ML demo to understand a model and its predictions in real time.

Problem formulation, multi-step strategies, walk-forward evaluation, and strong baselines.

Kerckhoffs' principle, threat models, and security intuition: why the key matters more than the algorithm.

Project demos, technical explanations, and short conceptual talks.

Academic poster: data visualization, analysis, and key results.

📍 Québec, Canada