These visualizations are built as interactive learning tools, combining machine learning concepts with real-time exploration. Each demo is designed to make abstract ideas easier to understand by showing how algorithms behave dynamically.
K-Means vs DBSCAN
Interactive comparison of centroid-based and density-based clustering, demonstrating how different algorithms handle non-linear structures, cluster density, and noise.
Gradient Descent Optimizers
Interactive exploration of SGD, Momentum, RMSProp, and Adam, visualizing how different optimization strategies converge across complex loss landscapes.
Neural Network Backpropagation
Step-by-step visualization of how neural networks learn through forward and backward propagation, including gradient flow, loss computation, and weight updates in real time.