Spinal Normality

Classification can be a very powerful machine learning technique. Classifying spine attributes with copious details can overwhelm people and in this case medical professionals.

This project aims to evaluate di erent types of classification machine learning techniques on spine data to better understand which methods yield more desirable results. In this project there are factors measured from x rays and MRIs that are classified as “Normal” or “Abnormal” spine. Several di erent types of models that are used are: logistic regressions, KNN and Random Trees. Furthermore, GridSearch is used to manipulate parameters, along with various cross validated tests.

Complete write up and code available on my GitHub.

• GitHub