HAutoML

HAutoML is an open-source Automated Machine Learning (AutoML) platform developed by OptiVisionLab, School of Information and Communications Technology, Hanoi University of Industry.

The platform automates the entire machine learning workflow — from data preprocessing, model selection, hyperparameter tuning, to model deployment — enabling users to easily build high-accuracy models without requiring in-depth expertise in programming or data science.

Want to read the verbatim Markdown documentation as on MkDocs?

🎯 HAutoML Vision

HAutoML is designed to democratize machine learning — enabling anyone (students, business analysts, domain experts) to build effective machine learning models without needing to master complex low-level engineering details.

✨ Key Features

🔬 Scientific Approach

1. Intelligent Preprocessing

Automatic feature type detection (numeric, categorical, text)
Adaptive missing value imputation (median / mode)
StandardScaler & OneHotEncoder normalization
ColumnTransformer pipeline preventing data leakage

2. Hyperparameter Tuning

Grid Search & Random Search
Bayesian Optimization (BO) & Genetic Algorithm (GA)
$k$-fold cross-validation ($k=5$)
Balancing accuracy vs compute efficiency

3. Multi-Criteria Evaluation

Classification: Accuracy, Precision, Recall, F1, Balanced Acc
Regression: MAE, MSE, RMSE, $R^2$ Score
Generalization validation against overfitting
Automated selection of champion model

Quickstart Guide

The fastest and most complete way to run Next.js, FastAPI, Worker cluster, MongoDB, MinIO, and Kafka:

BASH

# 1. Clone repository
git clone https://github.com/optivisionlab/AutoML.git
cd AutoML

# 2. Configure environment from templates
cp src/backend/temp.env src/backend/.env
cp src/frontend/temp.env src/frontend/.env

# 3. Launch full stack in background
docker-compose up -d --build

# To stop the system
docker-compose down

Microservices Architecture

┌─────────────────────────────────────────────────────────────┐
│                     Frontend (Next.js)                       │
│                  Web User Interface / Giao diện              │
└────────────────────────────┬────────────────────────────────┘
                             │ HTTP/REST
┌────────────────────────────▼────────────────────────────────┐
│                  Backend (FastAPI)                          │
│              API, Business Logic, Orchestration              │
└─┬────────────┬──────────────┬─────────────┬─────────────────┘
  │            │              │             │
  │            │              │             │
┌─▼─┐    ┌────▼────┐    ┌────▼─────┐  ┌──▼───────┐
│   │    │ MongoDB │    │  Apache  │  │  MinIO   │
│   │    │(Database│    │  Kafka   │  │ (Storage │
│   │    │ Metadata│    │ (Queue)  │  │ Datasets)│
│   │    └─────────┘    └─────┬────┘  └──────────┘
│   │                         │
│   └─────────────────────────┼─────────────────────┐
│                             │ Task dispatch       │
└─────────────────────────────┼─────────────────────┘
                              │
                   ┌──────────▼──────────┐
                   │       Workers       │
                   │  (Process Training  │
                   │   Jobs from Kafka)  │
                   └─────────────────────┘

📋 Recent Release History

Version	Date	Highlights	Status
v2.2.0	11/04/2026	Distributed Computing, Smart Scheduling, Account Classification	Latest
v2.1.0	05/12/2025	Optima Search (GA, Bayesian Optimization)	Supported
v2.0.0	19/10/2025	MapReduce 1.0, MinIO Object Storage, Async API	Stable (ISINC 2026)
v1.1.2	21/06/2025	UI Updates + Security Fixes	Supported
v1.1.0	03/06/2025	Model Deployment & Inference API	Supported
v1.0.0	17/05/2025	Initial Release - Nền tảng HAutoML	Initial Release

👥 Contributors

GitHub Contributors

We sincerely thank all contributors to the HAutoML project at OptiVisionLab:

Welcome to HAutoML

✨ Key Features

Dataset Management

Automated AutoML Pipeline

Deployment & Inference

Microservices Architecture

Backend API Docs

Scientific Publication

🔬 Scientific Approach

1. Intelligent Preprocessing

2. Hyperparameter Tuning

3. Multi-Criteria Evaluation

Quickstart Guide

Microservices Architecture

📋 Recent Release History

👥 Contributors

HAutoML: Open-Source for Automated Machine Learning