Depression has been increasing even before COVID-19. As a tool to aid in treatment of this increase,psychedelics have seen a resurrence in recent years due to clinical trials and breakthrough studies. psychedelics have been used the "betterment of well people" , PTSD, etc.
Given patient features and psychedelic drug use, can we create a model to predict depression to enable clinical improvement?
Overall the SVC model has the best performance on the recall(because we want to reduce false negatives) and use MCC score which is a more balanced metric for an unbalanced data.
Test recall: 0.771
MCC: 0.516
