Discretization of Gradient Flow Overview This repository explores and compares three gradient descent approaches— Explicit GD, Implicit GD, and Newton GD —using only NumPy for maximum transparency and ...

Abstract: This paper introduces a novel hybrid Particle Swarm Optimization (PSO) algorithm called the Gradient Enhanced PSO (GE-PSO) that selectively incorporates a Gradient Descent (GD) technique to ...

Abstract: This paper proposes a multi-UAV path planning method based on an improved Hummingbird Algorithm (AHA). To address the original AHA’s tendency to become trapped in local optima and its slow ...

A mid-level neural network implementation from scratch on the MNIST Zip-Digits dataset. Includes forward and backward propagation, adaptive activation function selection, training/validation split and ...