Applied Statistics and Machine Learning

At the end of the course, the student is able to:

1. Understand basics of Python, develop scripts using Jupyter / Google Colab
2. Organise, clean and analyze data using Numpy, Pandas and Matplotlib
3. Build predictive models for tabular data using Scikit Learn
4. Detect over/underfitting, apply regularization, and validate models
5. Choose the right modelling approach for a given data set

The availability of large data sets and sophisticated algorithms have opened many possibilities for "machine learning" (ML) - the automatic discovery of patterns in data. Computational tools have matured to the point that many of these algorithms are readily accessible. This course offers students an introduction to the tools of the trade, built around the Python programming language. The similarities and differences with classical statistics will be discussed to provide the necessary grounding, but the focus is on building hands-on skills to successfully build predictive models for large tabular data sets.

Students will be trained to:
• Use Python in a notebook (Jupyter/Google Colab) environment
• Assess whether a given modeling approach is appropriate
• Organise, clean, and visualise large data sets
• Build predictive models for tabular data, and assess their quality and validity

During the course several real-life datasets will be analysed, including one involving disease progression of intensive care patients, and a large dataset of heart valve measurements.

Variabel

Mandatory presence: attendance during the practicals and tutorials is mandatory.