Publications & patents.

Neural network predicting geomagnetic storm intensity from solar wind and Dst data via a quadratic neuron architecture, with optional intensity-level classification.

A practical framework for genuine, lasting discipline — crushing distractions and achieving goals, backed by two years of research.

How robotics and AI are transforming space exploration and discovery.

A comprehensive study of ML and DL techniques for malware classification.

A deep-learning feature-engineering strategy for exoplanet prediction and false-positive reduction.

Compares LSTM, CNN, Autoencoder+LSTM, and Transformer for Li-ion battery RUL on the CALCE dataset; hybrid models outperform standalone baselines.