MLJAR AutoML

We’re on our way to revolutionizing the field of machine learning and data analysis. Whether you're a small startup or a large enterprise, MLJAR is here to support you on your machine learning journey.

Data Science Tools

What is AutoML?

The most advanced Machine Learning tool with the simplest User Interface

A The mljar-supervised is an Automated Machine Learning Python package that works with tabular data. It’s a complete package that abstracts the common way to preprocess the data, construct the machine learning models, and perform hyper-parameters tuning to find the best model.-of-the-art tool designed for Data Science professionals. Our platform provides advanced automated machine learning capabilities, enabling effective model creation and data analysis process optimization.

Open-source MIT license

The Over 100 citations in research papers

More than 1 million downloads

Works with Python with versions 3.8, 3.9, 3.10, 3.11

Uniqe features: automated documentation, fairness metric

Variety of algorithms as: ensemble, stacked ensemble, XGBoost, CatBoost and more

Features

Discover the unique features that make our AutoML the state of the art. Find those that make your ML project advanced, super easy, and understandable.

Complete pipeline

Simplifies the entire machine learning process from data preprocessing to model deployment.

Golden feature

Efficiently pinpoints the most influential variables for optimal model performance.

Model leaderboard

Enables easy comparison and selection of models based on performance metrics.

Automated reports

The report from running AutoML will contain the table with information about each model score and the time needed to train the model.

Feature selection

MLJAR AutoML takes care of features preprocessing like missing values imputation and converting categoricals, it can also handle target values preprocessing.

Auto-saving models

All models in the AutoML are saved and loaded automatically. No need to call save() or load().

Hyperparameter tuning

Optimizes model performance by automatically searching for the best combination of hyperparameters, saving time and effort.

Variety of algorithms

Choose from many algorithms such as: XGBoost, CatBoost, Neural Networks, Decision trees, Random forest and many more...

Modes

Your comfort using our product is top priority. That is why we created 4 modes instead making you change settings all the time.

Explain

• Ideal for initial data analysis
• 75%/25% train/test split
• Explanations

Perform

• Production-ready ML pipeline
• 5 fold cross-validation
• Feature engineering
• Search for a model under constraint for prediction time on a single sample

Compete

• ML competitions under time budget
• Adjusted validation
• Train/test on 5 or 10 fold cross-validation
• Feature engineering
• Try many models

Optuna

• 10 fold cross validation
• Tune algorithm with Optuna Framework

Discover more.

Find details on our GitHub and in our documentation.