This repository explores the use of autoencoders for breast cancer detection using ultrasound image data.
The data used in this study is the Breast Ultrasound Images dataset from Kaggle: Link to dataset on Kaggle.
Three different autoencoder models are implemented:
-
Linear Autoencoder: A basic autoencoder with Tanh activation functions and dense layers as encoder and decoder.
-
Convolutional Autoencoder (CNN Autoencoder): A convolutional autoencoder that utilizes convolutional and pooling layers to extract features from the image data.
Feel free to use the notebook and contribute to the project!
I encourage you to explore the code, make modifications, and share your findings.