Scientific activity

The Laboratory looks for, develops, and improves analysis methods for large volumes of data. These methods allow for new high-quality results to be achieved for a given subject field. In particular, the Laboratory is interested in carrying out joint research with international organizations in the following areas:

• Developing methods to uncover outliers;
• Developing and testing distributed architectures for predictive models construction;
• Building predictive models under conditions of high-noise data;
• Developing methods for building predictors that work under time constraints;
• Developing unsupervised learning methods;
• Developing and approving methodologies for distributed data analysis workflow;
• Teaching skills for applying big data analysis methods and technologies;
• Developing algorithms for the safe transmission of big data.

To find scientific problems and data sets, the Laboratory is going to cooperate with Yandex, CERN, and different HSE divisions.

Disseminating the Laboratory’s Results

The results of the work the Laboratory carries out are published in leading international journals such as the Journal of Instrumentation and Physics Review Letters, and are presented at conferences like PyDataNIPS, KDD, ICML, CHEP, and PyData.