EduRisk · System Workflow

Student Data Collection

Multiple data sources aggregated in real-time

📊 Academic Records

📱 IoT Sensors

📝 Attendance Logs

Data Processing & Feature Engineering

Cleaning, normalization, and feature extraction

Data Cleaning

Normalization

Feature Extraction

Encoding

Machine Learning Model

Ensemble models for risk prediction

Random Forest

XGBoost

LSTM Neural Network

Risk Detection

Classification into risk categories

Low Risk

Medium Risk

High Risk

Recommendation Engine

Personalized interventions for each student

Study Resources

Tutor Matching

Faculty Dashboard

Real-time visualization and alerts

Risk Heatmaps

Student Profiles

Intervention Tracker

Detailed Process Flow

1. Data Ingestion

Real-time data collection from academic databases, IoT devices, and manual entries.

2. Preprocessing

Handling missing values, outlier detection, and feature scaling.

3. Feature Engineering

Creating derived features like trend indicators, engagement scores, and risk indicators.

4. Model Inference

Ensemble model prediction with confidence scores and explainability metrics.

5. Recommendation Generation

Rule-based + ML recommendations tailored to risk factors.

6. Alert & Visualization

Push notifications to faculty and updated dashboard views.

System Workflow