Comparative Analysis of Image Classification Algorithms: A Study of CNN, Logistic Regression, K-means, GAN, and SNN

Description

This project provides an experimental framework for evaluating different types of artifical intelligence models on cats and dogs image classification task. The implemented models are:

• Logistic Regression (Supervised ML)
• K-Means Clustering (Unsupervised ML)
• Convolutional Neural Network (Supervised DL)
• GAN Discriminator (Unsupervised DL)
• Spiking Neural Network

Getting Started

Dependencies

The project relies on several popular libraries. Make sure you have the following installed:

• Python 3.7.4
• NumPy (1.21.6)
• Matplotlib (3.5.3)
• Pillow (9.5.0)
• scikit-learn (1.0.2)
• TensorFlow (2.10.1)
• Torch (1.13.1)
• TorchVision (0.14.1)
• nengo (3.2.0)
• nengo-dl (3.6.0)

Installing

Install these dependencies using pip

pip install numpy==1.21.6 matplotlib==3.5.3 pillow==9.5.0 scikit-learn==1.0.2 tensorflow==2.10.1 torch==1.13.1 torchvision==0.14.1 nengo==3.2.0 nengo-dl==3.6.0

Usage

1. Prepare dataset Organize your dataset into the following format:

dataset/
  ├── train/
  │    ├── class1/
  │    ├── class2/ 
  └── test/
       ├── class1/
       ├── class2/
      

2. Run the script

pyhton main.py

Authors

Contributors names and contact info

• Janok N. Dinçer
• Çağan Çakır