My name is Sole, the leading instructor at Train in Data and the maintainer of Feature-engine, and together with a group of passionate data scientists and software developers, we maintain and expand the functionality of this Python library for feature engineering and feature selection for machine learning and its documentation, so you can better prepare your data to craft more powerful and interpretable machine learning models.
At Train in Data, we create intermediate and advanced online courses on machine learning, data science and AI software development, to help you boost your data science skills and leverage the power of this and other popular Python libraries, to create faster and robust machine learning pipelines.
We talk, blog and participate in podcasts about machine learning, software development and open-source, so you'll hear about us a lot on the digital sphere ;)
Check out the courses that we teach.
| Courses | What you will learn |
|---|---|
| Feature engineering for machine learning | Learn to create new features, impute missing data, encode categorical variables, transform and discretize features and much more. |
| Feature selection for machine learning | Learn to select features using wrapper, filter, embedded and hybrid methods, and build simpler and reliable models. |
| Master Hyperparameter Optimization for Tabular Learning | Learn about grid and random search, Bayesian Optimization, Multi-fidelity models, Optuna, Hyperopt, Scikit-Optimize and more. |
| Machine learning with imbalanced data | Learn about under- and over-sampling, ensemble and cost-sensitive methods and improve the performance of models trained on imbalanced data. |
| Feature engineering for time series forecasting | Learn to create lag and window features, impute data in time series, encode categorical variables and much more, specifically for forecasting. |
| Forecasting with Machine Learning | Learn to perform time series forecasting with machine learning models like linear regression, random forests and xgboost. |
| Machine Learning Interpretability | Learn to interpret the predictions of your white box and black box machine learning models. |
| Clustering and Dimensionality Reduction | Learn to extract information from unlabelled data through clustering and dimensionality reduction techniques. |
Find out more about machine learning through our books, and have the code at your fingertips.
| Books | Summary |
|---|---|
| Python feature engineering Cookbook, third edition | Over 70 Python recipes to implement feature engineering in tabular, transactional, time series and text data. |
| Feature selection in machine learning, second edition | Over 20 methods to select the most predictive features and build simpler, faster, and more reliable machine learning models. |
| Imbalanced Data: Myths, Mistakes and Modern Solutions | A critical look at class imbalance, moving beyond SMOTE and default thresholds toward cost-sensitive learning, proper threshold tuning, and evaluation metrics that reflect real-world requirements. |
The open-source libraries I contribute to.
| Library | About |
|---|---|
| Feature-engine | Multiple transformers for missing data imputation, categorical encoding, variable transformation and discretization, feature creation and more. |
Feature-engine is built and maintained by Soledad Galli together with a growing community of open-source contributors. Get to know everyone who's helped shape it:
Follow us on social media or through our website to be up to date with our latest news.
| Media | Summary |
|---|---|
| Train in Data | Enroll in our courses and books |
| I talk about data science, machine learning and how to become a data scientist. | |
| BlueSky | I post about data science, machine learning and how to become a data scientist. |
| Newsletter | I talk about data science, machine learning and how to become a data scientist. |
| Blog | I write about data science, machine learning, feature engineering and selection and more. |
We hope to see you around.