Identifying Cancer Types in Microscope Images of Lung and Colon Cells

This study explores the use of a Vision Transformer (ViT) model to classify diabetic retinopathy severity from fundus images, demonstrating promising accuracy through transfer learning and data augmentation techniques. While results fell short of state-of-the-art benchmarks, the model shows potential for improving early diagnosis through AI-driven medical imaging.

The goal of this research project is to predict whether or not a patient has skin cancer through a machine learning model that is developed from an image dataset. Skin cancer is extremely dangerous, as over 9500 people in the US are diagnosed with it daily. If detected early, patients will have a more likely chance of survival. I tested an MLP Classifier, Decision Tree Regressor, a Logistic Regression Model, and a KNN Model to compare various results and ultimately determine the best accuracy. The MLP Classifier had a 74.5% accuracy, the Decision Tree Regressor had a 74.1% accuracy, the Logistic Regression Model had a 68.8% accuracy, and the KNN Model had a 74.6% accuracy (all testing). We can see that the MLP Classifier, Decision Tree Regressor, and the KNN Model had around the same accuracy while outperforming the Logistic Regression Model. However, when comparing training data, there seems to be a large overfitting problem with most of the models.

This research investigates how the choice of training population, personalized versus more general datasets, impacts the performance and generalizability of machine learning models for predicting Coronary Artery Disease risk. By comparing models trained on datasets of varying specificity using data from the UC Irvine Machine Learning Repository, the study explores trade offs between accuracy and broad applicability to improve strategies for early CAD risk assessment and prevention.