2/1/2024 0 Comments Data dredging meaingOne of the problems, in this case, is the risk of overstating subgroup effects, intentionally or unintentionally. This is particularly important in the case of precision medicine, in which different subgroups may respond differently to new drugs or experimental therapies. One of the features that big data often exhibit, and that is currently only partially understood, is data heterogeneity, that is the fact that different subpopulations from which the data are collected often behave or react differently to specific interventions. Big Data is nowadays a buzzword with profound ramifications in fields ranging from medicine, learning, social surveillance and e-commerce. The ability to analyse and extract useful information from sets of structured or unstructured data that are too large or too complex to be dealt with using standard data-processing techniques is a crucial and very sought after goal in the Information Era, particularly after the diffusion of internet technologies. Professor Xuming He at the University of Michigan shows how a better understanding of subgroup selection in the big data era is necessary for providing valid statistical analysis to aid decision making under uncertainty. Although it is relatively easy to find impressive-looking associations in big data, for instance through the use of data mining, these associations can be spurious. They directly impact several disciplines, including precision medicine and individualised learning, which rely on information concerning individuals and groups to make predictions on the expected effects of a specific intervention on population subgroups. Statistical models are routinely used to derive inferences from large amounts of data.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |