State estimation, adaptive filtering, and machine-learning projects in MATLAB and Python.
Stack: MATLAB · Python · Jupyter
Kalman filter tracking a target's 2D state from noisy measurements, estimating position, velocity and acceleration through the predict/update cycle. Core state-estimation and sensor-fusion technique.
Optimal linear filtering to recover a signal from additive noise, using a Wiener filter derived from the signal and noise statistics.
A neural network built from scratch to solve XOR — the classic demonstration that a non-linearly-separable problem needs a hidden layer.
Regression on tabular housing data, covering preprocessing, feature handling, model fitting and evaluation.