The Predictive Society and Data Analytics Lab pursues
innovative research in data science. That means we are creatively
developing and applying high-dimensional network-based methods in
machine learning, statistics and artificial intelligence that can be
used for knowledge extraction from data.
Examples for data we are analyzing are high-throughput data from genomic experiments. For instance, the data can come from gene expression, methylation or RNA-seq experiments, e.g., generated from next-generation sequencing (NGS) technologies. In this context, we are mainly interested in biomedical and health related questions of complex disorders and their relation to basic biological mechanisms. We have experience working with data from many cancer types, including, breast cancer, colon cancer, bladder cancer, prostate cancer, kidney cancer and lymphoma but we are also involved in studies of cystic fibrosis, asthma and diabetes. In addition, we are working on text mining problems, e.g., from electronic health records (EHR) of patients, for automizing clinical prediction tasks. All of these studies are conducted in close collaboration with biologists and clinicians.
Other types of data we are analyzing come from the stock market, social media, social sciences, marketing, online gaming or business and industrial processes. We are particularly interested in studying ways to forecast or predict the system behavior, like gene functioning, consumer behavior or human personality, and utilize for this network-based approaches in combination with a statistical framework. Examples for indistrial partners are Nokia (Finland), Cargotec (Finland), Elisa (Finland), and Hughes insurance (UK).
For a very brief overview and an introductory video of our research see here.
Our research interests are in data science including:
• Machine Learning/Artificial Intelligence
• Computational Biology
• Network Science
• Text data science
Group members of our team:
• Postdocs: 1
• PhD students: 3
• MSc students: 13
• BSc students: 1