Artificial Intelligence & Machine Learning
Build AI models for classification and analysis of scientific data.
Overview
Go beyond classic analysis: learn to build, train and evaluate machine learning and AI models, with a special focus on biological data and medical imaging (in a research setting).
Learning objectives
- ✓ Understand the main families of machine learning algorithms
- ✓ Prepare and structure data for training
- ✓ Train and evaluate classification and regression models
- ✓ Discover deep learning for image analysis
- ✓ Interpret results and their limits within an ethical framework
Target audience
Researchers and data analysts wanting to integrate AI into their work.
Prerequisites
Python basics required (level of the "Python for Data Science" course).
Detailed program
- Supervised vs unsupervised learning
- Data preparation and splitting (train/test)
- Overfitting and cross-validation
- Classification (trees, forests, SVM)
- Regression and scoring
- Evaluation metrics (precision, recall, AUC)
- Neural networks: principles
- Image analysis (e.g. medical imaging) in research
- Tools: Keras / PyTorch (overview)
- Bias, interpretability and model limits
- Framing: a research tool, not a medical diagnosis
- Capstone project on an open dataset
Teaching methods
A mix of theory and hands-on exercises on real cases. Course materials provided.
Assessment
Continuous assessment through exercises and quizzes. Certificate of completion for each participant.
Funding
Eligible for funding by your employer, training fund or research institution. Quote on request.
Accessibility
Our courses are accessible to people with disabilities. Contact us to adapt the program.
⚠ A research tool, not a diagnosis
The models covered are research-assistance tools. They do not constitute a medical diagnosis.
Related courses
Python for Data Science
Master Python and the Anaconda ecosystem to analyze, visualize and model your data.
R language for statistical analysis
Analyze your data and produce publishable results with R and the tidyverse ecosystem.