MNIST-Digit-classification using Perceptron Learning algorithm

Overview

Recognised handwritten digits from MNIST Dataset by implementing perceptron learning algorithm

Methodology

Trained 10 perceptrons that as a group learned to classify the handwritten digits in the MNIST dataset. Each perceptron has 785 inputs and one output. Each perceptron’s target is one of the 10 digits, 0−9. The inputs for 785 consisits of 784 pixels representing 28 X 28 pixel image represented as gray scaled value 0-255 for single handwritten digit. The output of each perceptron is either 0 / 1 and each of these perceptrons learns using the perceptron learning algorithm