For this project, I have built a convolutional neural network to classify mushroom species from the Common genus's images Kaggle dataset. The network is based on a pretained ResNet50 network, finetuned with fast.ai v2.
This project includes:
- EDA of the initial Kaggle dataset
- Oulier analysis
- Image preprocessing
- Building and finetuning a pretrained ResNet50 convolutional neural network
- Model evaluation using confusion matrices and classification reports
- Python - The programming language for of this project
- fast.ai v2 - A deep learning library,
- Scikit-learn - Model evaluation,
- Pillow - Image processing,
- Seaborn and Matplotlib - Data visualization,