How shlokye16/psycsmmi works
PsycAI is a survey-based mental-health risk research project with a deployed Streamlit application. The user submits social-media usage and Likert responses; the app transforms responses into domain scores, runs a trained classifier, and presents a non-diagnostic result. The README also documents dataset analysis, model comparison, and model export. Those research stages are represented from documentation; their notebook internals were not sampled.
Main entry points: main.py Survey interface · fn.py Inference preprocessing · experiment.ipynb Exploratory analysis · final.ipynb Model training
Components 11
Research and training
- Survey dataset
dataset.csv - Exploratory analysis
experiment.ipynb - Model training
final.ipynb
Survey application
- Survey interface
main.py - Risk result
main.py
Inference pipeline
- Trained classifier
model.pkl - Inference preprocessing
fn.py - Weighted scoring
fn.py - Domain aggregation
fn.py - Risk prediction
fn.py
Other components
- Survey respondent
Connections 10
- Survey respondent → Survey interface: submits responses
main.py - Survey interface → Inference preprocessing: calls refine
main.py - Inference preprocessing → Weighted scoring: applies scoring
fn.py - Inference preprocessing → Domain aggregation: aggregates scores
fn.py - Survey interface → Risk prediction: calls predict
main.py - Risk prediction → Trained classifier: loads model
fn.py - Survey interface → Risk result: displays outcome
main.py - Survey dataset → Exploratory analysis: analyzed in
README.md - Exploratory analysis → Model training: informs modeling
README.md - Model training → Trained classifier: exports classifier
README.md