How manoj914822/mlproject works
This is a student math-score regression project with both a Flask prediction interface and model-training workflows. The web path turns form fields into a dataframe, loads saved preprocessing/model artifacts, predicts, and renders the result. Training is represented in two documented/code-supported forms: the standalone training pipeline and the ingestion-to-transformation-to-model-training component flow. The sampled code shows these paths are distinct; unsampled template details and any additional deployment wiring are not inferred.
Main entry points: app.py Flask routes · artifacts Model artifacts · data_ingestion.py Data ingestion · predict_pipeline.py Prediction pipeline
Components 15
Prediction web app
- Flask routes
app.py - Input form
templates/index.html - Result page
templates/home.html - Form dataframe
src/pipeline/predict_pipeline.py - Prediction pipeline
src/pipeline/predict_pipeline.py
Model training
- Data ingestion
src/components/data_ingestion.py - Student dataset
notepad/data/StudentsPerformance.csv - Train/test splits
artifacts/train.csv - Feature transformation
src/components/data_transformation.py - Model selection
src/components/model_trainer.py - Standalone training
src/pipeline/train_pipeline.py - Training CSV
artifacts/data.csv
Artifacts and utilities
- Model artifacts
artifacts - Model utilities
src/utils.py
Other components
- Web user
Connections 21
- Web user → Flask routes: submits form
- Flask routes → Input form: renders
- Flask routes → Form dataframe: builds dataframe
- Flask routes → Prediction pipeline: requests prediction
- Prediction pipeline → Model utilities: loads objects
- Model utilities → Model artifacts: reads artifacts
- Prediction pipeline → Model artifacts: transforms with
- Prediction pipeline → Result page: renders result
- Flask routes → Result page: renders
- Result page → Web user: shows score
- Data ingestion → Student dataset: reads
- Data ingestion → Train/test splits: writes splits
- Data ingestion → Feature transformation: invokes
- Feature transformation → Train/test splits: reads
- Feature transformation → Model artifacts: saves preprocessor
- Data ingestion → Model selection: invokes
- Model selection → Model utilities: evaluates models
- Model selection → Model artifacts: saves model
- Standalone training → Training CSV: reads
- Standalone training → Model utilities: saves objects
- Standalone training → Model artifacts: writes artifacts