Irina Belyaeva, Ph.D Email & Phone Number

Washington, Dc, Us

Washington, Dc, Us

Baltimore, Md, Us

Machine Learning in Healthcare

Baltimore, Md, Us

Research Scientist in Machine Learning for Signal Processing Lab in Department of Computer Science and Electrical Engineering (CSEE), develop Riemannian optimization methods and models for medical image 4D reconstruction and denoising using tensors and other high-dimensional data representations. Develop architecture for discovering and analyzing brain development patterns from magnetoencephalography (MEG) using high-powered statistical tensor decompositions. Design tensor computational methodology for developing multimodal data-driven fusion brain imaging data, including MEG and functional magnetic resonance imaging (fMRI). Create methods for fMRI spatiotemporal tensor decomposition using high temporal resolution.

Develop and implement multi-phased cloud data plans including implementation roadmaps. Design and initiate scalable data ingestion methods to transform multiple datasets. Assess, suggest, and choose technical options to resolve development and integration issues. Create real-time high-volume event-sourcing platform using distributed technologies including Kafka, DataStax Enterprise (NoSQL), Java, and Spark SQL. Design and development of online and offline machine learning algorithms for member behavioral learning. Invent and deploy real-time APIs and data connectors for data collection and use of member interactions from/for internal and external vendors. Promote visibility of measurable events include gaps in care, event-driven actions through APIs, and Member Analytical Platform. Devise various streaming data services for live master data management, enrichment, and building analytical views. Architected Kafka File Processing Engine to enter event data from external vendors.

La Jolla, Ca, Us

Arranged bioinformatics data mining workflows to construct federated NoSQL datastore from numerous related data sources (Cassandra, Oracle, Postgres, MySQL, flat files). Developed logical and physical data prototype of the subject-oriented integrated biological data warehouse to secure genomic, genetic, and phenotypic datasets, genome annotation results, and image/binary data (Postgres). Devised and executed production quality data analysis pipelines to transfer and process large datasets in parallel with datasets dynamic partitioning, and multithreading (Spark, Spring Batch on Hadoop/Java, Oracle, Postgres). Designed portal federated search (data content, and web content) featuring Elastic Search. Produced software design documents, client presentations. Liaised with domestic and international project stakeholders. ♦ Created diverse set of software components to make large-scale genomics data mining feasible. ♦ Produced Arabidopsis Information Portal (Araport) open-access online community resource for Arabidopsis research producing web-portal for bioinformaticians to perform online genomics research, exploratory data analysis, data mining, and genes browsing, and annotation of new release of Arabidopsis thaliana reference genome assemble, and scientific publications ThaleMine: A Warehouse for Arabidopsis Data Integration and Discovery.”

Planned and delivered an algorithmic framework for efficient satellite resource use and consumption resulting in early deployment of safety-critical software compoent by NASA team to decommission outdated software product.

College Park, Md, Us

Liaised with university Biology department researchers and international experts to develop Bioinformatics Ecological statistical modeling software package for Animal Movement Tracking initiative featuring statistical languages and data analysis methods, and data mining algorithms. Created high-level information architecture diagrams documenting system and data flows, core statistical modules using R, and application programming interfaces.

Ph.D, Computer Science, Machine Learning And Signal Processing

Master'S Degree, Computer Science