During the last years, the field of Machine Learning grew rapidly, mainly due to improvements in its algorithms, the increase of data availability and a reduction in computing costs. This growth is having a profound impact in biomedical applications, where the great variety of tasks and data types enables us to get benefit from Machine Learning algorithms in many different ways.
We use Machine Learning algorithms to solve biomedical problems with large datasets. We are particularly focused in the following applications:
- Structured time series: Temporal time series of structured data often appear in biomedical datasets, presenting challenges such as containing variables with different periodicities, being conditioned by static data, etc. We train algorithms to find complex patterns in these datasets and with them we build systems such as early warning systems and medical decision support systems.
- Genomics data: Machine Learning in genomics is still an emerging subfield, but given that genomics data are arguably the most extensive and complex datasets that can be found in biomedicine, it is expected that many relevant Machine Learning applications will arise in the near future. We develop Machine Learning algorithms specially tailored for the characteristics of genomic datasets.
- Medical notes: Vast amount of medical observations are stored in the form of free text, and in our group we develop techniques and make use state of the art algorithms to extract such information.
To tackle these problems we work with multiple types of Machine Learning models, such as Deep Learning models and probabilistic models. We apply state of the art algorithms and also develop new theoretical improvements specifically tailored to our tasks of interest.