Neural Networks (Dry Bean)

Introduction

This project was made for an assignment in the course Neural Networks, Artificial Intelligence and Big Data at the IMC University of Applied Sciences Krems.

Goal

The goal of this project was to create a neural network that predicts the class of the beans by using the other variables as predictors. The dataset used for this project is the Dry Bean Dataset provided by the University.

Tasks

• Load the Dry_Bean dataset (attached) and:
• Perform a short Explorative Data Analysis (EDA)
• Create a feed-forward neural network to predict the class of the beans by using the other variables as predictors
• Split the data in train and test (with ratio 80-20)
• Convert to tensors
• Experiment with different number of layers and nodes, different learning rates, different optimizers, etc.
• Show results on training and test set
• Due date: 22.05.2023, 09:00 CEST

---

Assignment Evaluation

Possible points for assignment: 10

Feedback: Excellent! You were one of the two students doing the rescaling correctly. Instead of removing 1 pts from all other students, I decided to add 1pt bonus to you two. Punkte 11 / 10

Report

The report can be found in the file report/nn.html or src/nn.ipynb.