Compared the accuracy and training time of two deep learning models (AlexNet and Densenet121) with that of an ensemble of traditional ML algorithms (KNN and Decision Tree). This was done to determine whether the ensemble could prove to be a viable approach for automating cervical spine fracture detection. Dataset taken from the 2022 RSNA Cervical Spine Fracture Detection competition on Kaggle.

Pretrained models were used. Projects done for experimental purposes.

AP Research paper for this project: https://drive.google.com/file/d/17OGHVGk_TAHMw1l6L3-yBiT7RiJt-jyo/view?usp=sharing