Collected 9000+ images to create a dataset. Applied Gaussian noise to make images noisy. Trained an autoencoder for 200 epochs to remove noise from the images. Achieved an accuracy of 79.52% Original Images After introducing noise to images Loss/Accuracy Graph Final Results Link to GitHub Repository

Created two binary class classification models using Support Vector Machine classifier and Artificial Neural Network to predict if tumor is malignant or benign. Used the Breast Cancer Wisconsin dataset with 569 instances. Applied GridSearchCV on SVM classifier to find the best parameters. Tuned the parameters of ANN Model to achieve better accuracy. Achieved accuracy of 97.37%, precision of 0.9565, recall of 1.0 and F1-score of 0.9777 using SVM model.