This is a job we received from the company "skinCare". The job entailed that we would develop a website that would be able to classify pictures of moles.
The dataset provided by the client is here https://www.kaggle.com/kmader/skin-cancer-mnist-ham10000
- Python 3.7
- Tensorflow
- sklearn
- pandas
- numpy
- matplotlib.pyplot
- seaborn
- plotly.express
- streamlit
The model was overfitting by a lot.
To improve accuracy there has to be looked at more data augmentation techniques
to filter the images given to the model.