In this class, you will have the opportunity of attending hands-on tutorials to move the lecture concepts to code and play with tools and functions you will use for implementing your project in the second half of the semester. To ensure you will be ready for starting the project, you will be asked to solve homework exercises that challenge you on applying the concepts explained in the lectures and the programming solution shown in the tutorials. For tutorials and homework, you will rely on the packages provided for data analysis and machine learning with Python (e.g., NumPy, SciPy, Pandas, Matplotlib, Seaborn, Bokeh, ScikitLearn, TensorFlow, PyTorch, ...).
The goal of lab sessions is for you to learn how to apply different concepts to the analysis of behavioral data and the implementation of machine-learning approaches suitable for predictive tasks. Through the tutorials, we aim to provide you with guidance on how to set up a programming environment and on which and how the most important packages can be used to deal with behavioral data. By the end of lab sessions, you should be able to:
- Setup a programming environment for running projects on behavioral data.
- Understand and apply programming tools and functions for data handling and visualization.
- Code experimental pipelines for classification and regression analyses.
- Implement neural networks and integrate them in experimental pipelines.
- Create experimental pipelines for analyses that require unsupervised learning techniques.
- Develop recommendation pipelines, from behavioral data to recommended lists.
- Implement techniques for learning latent representations from behavioral data.
- Applying techniques for combining multiple data sources in your experimental pipelines.
This code is provided for educational purposes and aims to facilitate reproduction of lab sessions. We have done our best to document, refactor, and test the code before publication.