Machine Learning Skin Disease Classification

Project Infographic:

As part of my data science class, I collected and preprocessed thousands of images sourced from DermNet of 23 skin diseases, then fine tuned a CNN (convolutional neural network) model to classify any image of a skin condition into 1 of 23 given categories. The final model achieved 40% accuracy on the test data.

To dive futher into this project, check out my source code or watch my video summary.