Skip to content

adityonugrohoid/telecom-ml-portfolio

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

Telecom AI/ML Portfolio

Portfolio Project: Demonstrating AI/ML application to real-world telecom challenges using domain expertise from 10+ years in network operations.

Python 3.11+ uv License: MIT


Projects

# Project ML Type Algorithm Target Result
1 Churn Prediction Binary Classification XGBoost is_churned AUROC: 0.86
2 Root Cause Analysis Multi-class Classification XGBoost is_root_cause Acc@1: 0.91
3 Anomaly Detection Unsupervised Isolation Forest label_anomaly F1: 0.70
4 QoE Prediction Regression LightGBM mos_score RMSE: 0.45
5 Capacity Forecasting Time-Series LightGBM+Prophet traffic_load_gb MAPE: 14.5%
6 Network Optimization Reinforcement Learning Q-Learning KPI improvement +61% vs random

What This Portfolio Demonstrates

  • Domain Expertise: Deep understanding of telecom network operations and challenges
  • Problem Framing: Translating business problems into well-defined ML tasks
  • Data Engineering: Hand-crafted synthetic data generators with embedded telecom physics
  • End-to-End Thinking: Data generation, feature engineering, modeling, evaluation, and business insights
  • Communication: Clear documentation for both technical and business audiences

Technology Stack

Category Tools
Language Python 3.11+
Package Manager uv
ML Frameworks XGBoost, LightGBM, scikit-learn
Data Processing Pandas, NumPy
Visualization Matplotlib, Seaborn
Interpretability SHAP
Testing pytest
Linting Ruff
CI/CD GitHub Actions

Quick Start

Each project is fully independent. To explore one:

cd 01-churn-prediction
uv sync
uv run python -m churn_prediction.data_generator
uv run jupyter lab notebooks/

See each project's QUICKSTART.md for detailed instructions.


Project Structure

Every project follows the same structure:

0X-project-name/
├── .github/workflows/ci.yml    # CI pipeline
├── data/
│   ├── raw/                    # Generated synthetic data
│   └── processed/              # Feature-engineered datasets
├── src/{package_name}/
│   ├── __init__.py             # Package exports
│   ├── config.py               # Configuration management
│   ├── data_generator.py       # Domain-informed data generation
│   ├── features.py             # Feature engineering pipeline
│   └── models.py               # ML model implementations
├── notebooks/
│   └── 0X_analysis.ipynb       # Main analysis notebook
├── tests/
│   └── test_data_quality.py    # Data quality tests
├── .gitignore
├── pyproject.toml
├── README.md
├── QUICKSTART.md
└── CONTRIBUTING.md

Author

Adityo Nugroho Telecom Professional | AI/ML Practitioner


License

All projects are MIT licensed for educational and portfolio purposes.

About

Telecom AI/ML Portfolio — 6 end-to-end ML projects demonstrating domain expertise in network operations

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors