https://github.com/vickshan001/fashion-mnist-cnn-classifier
Deep learning project using CNNs to classify Fashion MNIST images. Includes activation function comparison, dropout, and learning rate tuning.
https://github.com/vickshan001/fashion-mnist-cnn-classifier
artificial-intelligence cnn deep-learning fashion-mnist image-classification keras python tensorflow
Last synced: 2 months ago
JSON representation
Deep learning project using CNNs to classify Fashion MNIST images. Includes activation function comparison, dropout, and learning rate tuning.
- Host: GitHub
- URL: https://github.com/vickshan001/fashion-mnist-cnn-classifier
- Owner: vickshan001
- Created: 2025-03-30T21:39:00.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2025-03-30T21:41:09.000Z (6 months ago)
- Last Synced: 2025-03-30T22:27:22.787Z (6 months ago)
- Topics: artificial-intelligence, cnn, deep-learning, fashion-mnist, image-classification, keras, python, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 0 Bytes
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 🧥 Fashion MNIST CNN Classifier (2024)
The goal is to build, train, and evaluate a Convolutional Neural Network (CNN) to classify clothing items from the Fashion MNIST dataset.
---
## 🧠 Model Features
- 👕 Classifies images into 10 categories (e.g., T-shirt, Trouser, Sneaker)
- 🔍 Explores three activation functions: **Tanh**, **Sigmoid**, **ELU**
- ⚙️ Tests five learning rates: `0.001`, `0.1`, `0.5`, `1`, and `10`
- 📉 Uses **Categorical Cross-Entropy** as the loss function
- 🧪 Includes a **Dropout Layer** for regularization
---
## 📊 Summary of Findings
### 🔸 Activation Function Comparison
| Activation | Train Accuracy | Test Accuracy |
|------------|----------------|---------------|
| Tanh | 100% | 91.73% |
| Sigmoid | 87.87% | 87.04% |
| ELU | 98.41% | 91.25% |
### 🔸 Learning Rate Comparison
| Learning Rate | Train Accuracy | Test Accuracy |
|---------------|----------------|----------------|
| 0.001 | 84.17% | 83.11% |
| 0.1 | 98.84% | 90.80% |
| 0.5 – 10 | ~10% | ~10% |
---
## 📂 Files Included
- `fashion_mnist.ipynb` – CNN implementation and experimental results
- `AI_CW2_Report_Vickshan_Vicknakumaran.pdf` – Supporting coursework documentation
---
## 🏫 Info
- 🏫 University: Queen Mary University of London
- 📅 Year: 2024
- 👨💻 Author: Vickshan Vicknakumaran
---
## 🚀 How to Run
1. Clone the repository.
2. Open `fashion_mnist.ipynb` in Jupyter Notebook.
3. Make sure the following libraries are installed:
- `tensorflow`, `keras`, `numpy`, `matplotlib`
4. Run the notebook to train and evaluate the model.
---
## 📜 License
For academic and research use only.