首页|Ludwig-Maximilians-Universitat Munchen Reports Findings in Machine Learning (COD I: Enhancing machine learning-based molecular profiling through contextual out-o f-distribution integration)
Ludwig-Maximilians-Universitat Munchen Reports Findings in Machine Learning (COD I: Enhancing machine learning-based molecular profiling through contextual out-o f-distribution integration)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning DailyNews Daily News-New research on Machine Learning is th e subject of a report. According to newsreporting from Bavaria, Germany, by New sRx journalists, research stated, "Molecular analytics increasinglyutilize mach ine learning (ML) for predictive modeling based on data acquired through molecul ar profilingtechnologies. However, developing robust models that accurately cap ture physiological phenotypes ischallenged by the dynamics inherent to biologic al systems, variability stemming from analytical procedures,and the resource-in tensive nature of obtaining sufficiently representative datasets."Funders for this research include LMU Munich, Helmholtz Zentrum Munchen.The news correspondents obtained a quote from the research from Ludwig-Maximilia ns-UniversitatMunchen, "Here, we propose and evaluate a new method: Contextual Out-of-Distribution Integration(CODI). Based on experimental observations, CODI generates synthetic data that integrate unrepresentedsources of variation enco untered in real-world applications into a given molecular fingerprint dataset. B yaugmenting a dataset with out-of-distribution variance, CODI enables an ML mod el to better generalizeto samples beyond the seed training data, reducing the n eed for extensive experimental data collection.Using three independent longitud inal clinical studies and a case-control study, we demonstrate CODI'sapplicatio n to several classification tasks involving vibrational spectroscopy of human bl ood. We showcaseour approach's ability to enable personalized fingerprinting fo r multiyear longitudinal molecular monitoringand enhance the robustness of trai ned ML models for improved disease detection."
NETL
NSTL
