Synthetic intelligence (AI) has the potential to rework all areas of science, power, and safety. Neeraj Kumar, Chief Information Scientist within the Superior Computing, Arithmetic, and Information Division at Pacific Northwest Nationwide Laboratory (PNNL), advances biomedical research with AI.
As a part of the Most cancers Distributed Studying Setting (CANDLE) challenge, Kumar and his staff, in collaboration with Argonne Nationwide Laboratory, developed a way to enhance the AI-based prediction of drug responses throughout most cancers remedy. He introduced his outcomes on the Platform for Superior Scientific Computing (PASC) Convention on June 3 in Zurich, Switzerland.
In his speak, titled “Basis Fashions and Giant-Scale Workflow for Most cancers Evaluating Metrics for Machine studying Fashions,” Kumar described the Comparability Workflow for Cross Validation (CMP-CV) framework, which standardizes the coaching and analysis of a number of deep studying fashions.
“Variations in how deep studying fashions are crafted, educated, and examined make direct comparability tough,” mentioned Kumar. “The CMP-CV framework permits scientists to make this comparability, resulting in the event of higher fashions, corresponding to these that may precisely predict drug responses.”
Kumar additionally participated in a panel dialogue, the place he spoke concerning the makes use of of AI in personalised drugs to construct most cancers digital twins.
“The combination of AI, co-design, and computing helps scientists deal with grand challenges, like these in human well being,” mentioned Kumar. “In the way forward for personalised drugs, AI can create digital replicas of particular person sufferers to empower medical doctors to tailor therapies with precision, optimizing outcomes and reworking healthcare.”
Because the AI revolution continues to unfold, Kumar’s work underscores the immense potential for AI to rework scientific discovery on a world scale. By fostering collaborative analysis and transcending area boundaries, Kumar is on the forefront of the applied sciences accelerating the interpretation of scientific insights into real-world options.
His latest involvement in co-organizing the Trillion Parameter Consortium (TPC) European Workshop in Barcelona additional demonstrates Kumar’s dedication to advancing science and engineering by way of large-scale fashions, co-design, and cutting-edge computing programs. The workshop started with a tutorial on massive language fashions (LLMs), retrieval-augmented era (RAG), and AI brokers, the place Kumar supplied deep insights into the expansive functions and growth of LLMs inside the scientific neighborhood. PNNL atmospheric scientist Po-Lun Ma additionally participated on this workshop by giving a lightning speak on knowledge challenges from local weather science.
Kumar’s management extends past his analysis endeavors. He serves on the management staff of the Heart for AI @PNNL, which drives strategic initiatives and shapes the long run path of AI analysis on the laboratory. The Heart is an important element in supporting the Division of Power’s proposed Frontiers in Synthetic Intelligence for Science, Safety and Expertise (FASST) initiative, which can drive advances in AI-ready knowledge, frontier-scale compute, and secure, safe, and reliable fashions, thereby accelerating discovery in science, safety, and expertise.
The CANDLE challenge is supported by the Division of Power’s Exascale Computing Mission by way of the Superior Scientific Computing Analysis program.
EIN Presswire gives this information content material “as is” with out guarantee of any type. We don’t settle for any accountability or legal responsibility
for the accuracy, content material, pictures, movies, licenses, completeness, legality, or reliability of the data contained on this
article. When you’ve got any complaints or copyright points associated to this text, kindly contact the writer above.
