How tdiprima/applied-ml works
This repository is a collection of seven standalone applied-ML examples, not a single end-to-end service. Each script reads its own CSV, trains or fits a model, and presents task-specific predictions, evaluations, summaries, or forecast/anomaly artifacts. The graph preserves the documented classification, segmentation, forecasting, anomaly-detection, and regression projects. Source was sampled for all seven scripts; links show only their documented or directly visible data and execution flow.
Main entry points: train_model.py Lead qualification · classifier.py Ticket classifier · email_classifier.py Spam filter · segment_customers.py Customer segmentation
Components 16
Classification
- Support tickets
src/logreg-ticket-classifier/support_tickets.csv - Ticket classifier
src/logreg-ticket-classifier/classifier.py - Training leads
src/rf-lead-qualification/leads_train.csv - Test leads
src/rf-lead-qualification/leads_test.csv - Lead qualification
src/rf-lead-qualification/train_model.py - Email examples
src/nb-spam-filter/data/emails.csv - Spam filter
src/nb-spam-filter/email_classifier.py
Segmentation
- Customer records
src/kmeans-customer-segmentation/customers.csv - Customer segmentation
src/kmeans-customer-segmentation/segment_customers.py
Forecasting
- Task history
src/prophet-task-forecaster/task_volumes.csv - Task volume forecast
src/prophet-task-forecaster/prophet_forecaster.py - Revenue history
src/lr-revenue-forecaster/data/revenue_data.csv - Revenue prediction
src/lr-revenue-forecaster/revenue_predictor.py
Prediction
- System metrics
src/isolation-forest-system-anomaly/system_metrics.csv - Anomaly detection
src/isolation-forest-system-anomaly/anomaly_detector.py
Other components
- Analyst
Connections 22
- Analyst → Ticket classifier: runs
- Support tickets → Ticket classifier: loads
- Ticket classifier → Analyst: classifies tickets
- Analyst → Lead qualification: runs
- Training leads → Lead qualification: loads training data
- Test leads → Lead qualification: loads test data
- Lead qualification → Analyst: reports lead scores
- Analyst → Spam filter: runs
- Email examples → Spam filter: loads
- Spam filter → Analyst: classifies email
- Analyst → Customer segmentation: runs
- Customer records → Customer segmentation: loads
- Customer segmentation → Analyst: reports segments
- Analyst → Task volume forecast: runs
- Task history → Task volume forecast: loads
- Task volume forecast → Analyst: reports forecast
- Analyst → Revenue prediction: runs
- Revenue history → Revenue prediction: loads
- Revenue prediction → Analyst: reports predictions
- Analyst → Anomaly detection: runs
- System metrics → Anomaly detection: loads
- Anomaly detection → Analyst: reports anomalies