### Course Description

A Big Data revolution is currently under way in Health Care and Clinical Medicine. The availability of rich biomedical data, including Electronic Health Records, -Omics Data, and Medical Imaging Data, is driving Precision Medicine, which is a recent approach for disease treatment and prevention tailored to the specifics of the individual patient. Precision Medicine is expected to result in significant benefits for both patients and practitioners. Therefore, new approaches that can make sense of complex Biomedical Data and provide data-driven insights are in high demand. This course provides an introduction about current data-related problems in the field of Clinical Medicine and overviews how Machine Learning and Data Mining can leverage complex biomedical data. The course includes a series of lectures covering many basic concepts, as well as a panel of hands-on demonstration that will help you getting started with R and some of the commonly used R libraries to prepare and analyze the data. You should enroll if you are a Computer Scientist, a Data Scientist, or a professional in the Biomedical field with some hands-on experience in Computer Programming (a basic knowledge of R is preferred, even if not required) and you want to learn how to tackle Data Mining problems in the field of Clinical Medicine using R and Machine Learning.

#### What am I going to get from this course?

**Learn to use R to leverage complex biomedical data by Machine Learning approaches.**Students will learn the

**basic concepts of Data Mining, Machine Learning and the specifics of the different data types generated in the context of modern Clinical Medicine**. Also, a series of hands-on demonstration will help getting started with Machine Learning-assisted Clinical Medicine using widely used R libraries.

- Questions and data types in the medical field, including Electronic Health Records, Medical Imaging Data, and -Omics Data
- CRISP-DM method applied to Clinical Medicine problems
- Data Preparation using R
- Non-negative Matrix Factorization (NMF) as effective tool for feature extraction
- Feature selection, including regularization methods (LASSO, Elastic Net)
- Building classification and numeric prediction models using R (regression, trees, ensemble methods, neural networks, and others)
- An example of deep learning using R and H2O
- Estimating model performance and accounting for error
- Iterative optimization to find a local error minimum: an example of gradient descent
- How to Deploy Predictive Models in the Medical Field