Machine learning for psychotic disorders clinical trials

Keywords: Classification, Clinical trials, Predictive modeling, Data visualization, Data mining

1. Performed feature selection across 9 machine learning models and statistical tests to comprehensively assess and identify the most relevant predictors in Schizophrenia diagnosis

2. Optimized classifiers and feature selection models by fine-tuning hyper-parameters with randomizedSearchCV and implementing cross-validation techniques

3. Achieved ~80% accuracy with the optimized random forest model outperforming KNN and logistic regression models

Python Jupyter Notebook Scikit-Learn Matplotlib H2O AutoML Statsmodel