Popular repositories Loading

1. Neural-Network-for-Handwritten-Digit-Recognition-MNIST---83-Accuracy Neural-Network-for-Handwritten-Digit-Recognition-MNIST---83-Accuracy Public

   A two-layer neural network built from scratch only using numpy and simple mathematics that achieves ~83% accuracy on handwritten digit recognition using the MNIST dataset. The implementation includ…

   Jupyter Notebook

2. EdgeHAR-Efficient-Human-Activity-Recognition-Using-IMU-Sensors EdgeHAR-Efficient-Human-Activity-Recognition-Using-IMU-Sensors Public

   This project addresses the challenge of deploying Human Activity Recognition (HAR) on edge devices with limited processing power. We processed 50Hz accelerometer data from 2 three-axial Axivity AX3…

   MATLAB