Machine learning in Python with scikit-learn

The workshop will take place at Science Park 402, 1098 XH Amsterdam. Please note that lunch and drinks at the end of the workshop are included.

This hands-on workshop will provide you with the basics of machine learning using Python.

Machine learning is the field devoted to methods and algorithms that ‘learn’ from data. It can be applied to a vast range of different domains, from linguistics to physics and from medical imaging to history.

This workshop covers the basics of machine learning in a practical and hands-on manner, so that upon completion, you will be able to train your first machine learning models and understand what next steps to take to improve them.

We start with data exploration so that it is suitable for machine learning. Then we learn how to train a model on the data using scikit-learn. We learn how to select the best model from the trained models and how to use different machine learning models (like linear regression, logistic regression, decision tree and support vector machine), and discuss some of the best practices when starting your own machine learning project.

The workshop is based on the teaching style of the Carpentries, and learners will follow along while the instructors write the code on screen. More information can be found on the workshop website.

Who

The workshop is open and free to all researchers in the Netherlands at PhD candidate level and higher. We do not accept registrations by Master students. The workshop is aimed at PhD candidates and other researchers or research software engineers.

The course aims to be accessible without a strong technical background.

This course is for you if:

• You have basic knowledge of Python programming : defining variables, writing functions, importing modules. Some prior experience with the NumPy, pandas and Matplotlib libraries is recommended but not required.
• You want to learn how to setup a full machine learning pipeline in Python for various machine learning tasks.
• You want to get an intuition of basic machine learning concepts, such as train-test data splits, model training and evaluation, different machine learning algorithms, overfitting/underfitting, bias-variance trade-off.

This course is not for you if:

• You already have experience with machine learning or its concepts, this is really an introduction for people that have never done machine learning or only just started but need more guidance.
• You want to get a solid mathematical understanding of machine learning theory. This course aims to quickly get participants comfortable applying machine learning in practice, we therefore only cover the basis of theoretical concepts without going into depth.
• You want to learn about deep learning
• You want to learn about more advanced data preprocessing, like data cleaning, handling missing values etcetera. We only cover the basics of data preprocessing that are needed to setup a machine learning pipeline.

Also have a look at the syllabus to see what topics we will cover.

If you are uncertain whether this course is for you, please send us an email.

Syllabus

Machine learning concepts

• What is machine learning?
• Different types of machine learning
• Big picture of machine learning models
• General pipeline

The predictive modeling pipeline

• Tabular data exploration
• Fitting a scikit-learn model on numerical data
• Handling categorical data

Selecting the best model

• Overfitting and underfitting
• Validation and learning curves
• Bias versus variance trade-off

Machine learning best practices

• Data hygiene
• Correct evaluation
• How to keep your machine learning project organised

Where

This training will take place in-person at the eScience Center, Science Park 402, Amsterdam.