Neural Network Implementation for MultiClass Classification in Python

Overview

This repository contains the Implementation of a neural network for multiclass classification in Python from scratch. Two distinct datasets were utilized to demonstrate the functionality of the neural network.

Datasets

1. Random People Dataset:

   • This dataset comprises information on height, weight, and gender of randomly selected individuals.

2. Glass Identification Dataset (UCI):

   • The UCI Glass Identification dataset consists of 10 attributes, including an identification column. The response variable is the type of glass, categorized into seven discrete values. The last column contains the labels for the glass types.

Project Tasks

1. Data Splitting:

   • The dataset was divided into two categories: training data (70%) and test data (30%).

2. Neural Network Creation:

   • A neural network was developed to handle the multi-class classification task.

3. Performance Metrics:

   • Evaluation of the neural network performance was conducted on the test data, reporting comprehensive classification metrics.