Requirements:
a. Good level of mathematical education (pure/applied mathematics); knowledge in topology/discrete geometry, graph theory, numerical analysis, statistics is a plus;
b. Existing knowledge of standard machine learning algorithms; experience of working with some of them is a plus;
c. Existing knowledge of standard algorithms in statistical data analysis; experience of working with data is a plus;
d. Programming skills (any language, preferably in Python/R; if the language is different from Python/R, then willingness to switch); experience with SciPy and similar scientific packages is a plus;
e. Creative, analytic and strategic thinking; experience of participation in R&D projects is a plus;
f. Presentation skills (to be able to convey the logic of your discoveries in the team);
g. Good command of English.
Tasks:
— development of new, and modification of the existing, machine learning algorithms tailored to specific tasks in clinical trials, with specific focus on TDA;
— programming implementation of algorithms;
— module integration and interaction with the legacy code;
— conducting experiments with available datasets.
Topology-based Clinical Data Mining (TCDM) is a novel methodology that combines Biostatistics, Topological Data Analysis, Machine Learning, and Data Visualization. Its goal is to discover hidden patterns within a set of interrelated clinical outcomes by extracting comprehensive “topological maps” of a dataset represented in the form of the graph. This geometric, data-driven approach was adopted to develop proprietary algorithms and, hence, a computational platform that can be used by researchers to perform data mining experiments on clinical datasets.