Anirudh

PrabhuHe/His

Petrology, Mineralogy & Mineral Physics Astrobiology & Geobiology

Anirudh Prabhu’s research involves developing algorithms, visualizations, and methods in the fields of artificial intelligence, machine learning, and informatics and their application in a variety of domains such as Earth and Space Science, Paleobiology, and Microbiology.

The path to achieving this goal lies in a truly multidisciplinary approach, harnessing the integrated power of multiple scientific domains by using data as the backbone of the discovery process. This in turn required solving problems that plague the fields of data science, knowledge extraction, and data visualization, for example, how to deal with extremely small or sparse data, how to extract the relationships embedded in unstructured text, and how to deal with scalability issues with an extremely large number of attributes for data mining algorithms, such as affinity analysis.

Prabhu’s Ph.D. thesis work focused on developing a semi-automated methodology to create rules by identifying the important concepts in text data and the relationships between these concepts. Prabhu was the Student Representative of the Geo-informatics and Data Science Division of the Geological Society of America from 2018 to 2021, and has been a member of the Data Science Team of both the Deep Carbon Observatory and the 4D initiative since 2016.

Recent advances in data generation techniques, whether by experiments, measurements, or computer simulation, quickly provide complex data characterized by source heterogeneity, multiple modalities, often high volume, high dimensionality, and multiple scales (temporal, spatial, and function). In turn, science and engineering disciplines are rapidly becoming more and more data-driven by a variety of goals (the Deep Carbon Observatory is an exemplar); higher sample throughput, high resolution, additional physics/ chemistry/ biology, new instrumentation, and new integrated databases all with the ultimate aim of better understanding/modeling of the complex systems and their dynamics that underlie the processes being studied. However, analyzing libraries of complex data requires managing the inherent complexity to allow integration of the information and knowledge across multiple scales and spanning traditional disciplinary boundaries. Significant advances in methods, tools, and applications for data science and informatics over the last five years can now be applied to multi- and inter-disciplinary problem areas. Virtual Observatories, Virtual Organizations, complex networks, linked data across systems, full life cycle data management, data integration, citation, and attribution are now increasingly becoming an integral part of projects whether small (few people, one organization, modest data needs) or the very large (many investigators, organizations, diverse data needs).

Given this increasing data deluge, it is clear that each of the Directorates in the Deep Carbon Observatory faces diverse data science and data management needs to fulfill both their decadal strategic objectives and their day-to-day tasks. This project will assess in detail the data science and data management needs for each DCO directorate and for the DCO as a whole, using a combination of informatics methods; use case development, requirements analysis, inventories, and interviews.

Learn more open_in_new