Debashis Ghosh is a biostatistician whose research includes recurrent-event methods, statistical machine learning and biomarker evaluation. The Colorado School of Public Health identifies him in its Department of Biostatistics and Informatics. His fuller name is also printed directly in the publisher’s byline for the 2000 paper with D. Y. Lin.[ref]
Debashis Ghosh profile at the Colorado School of Public Health. Institutional directory, research description
[ref]That joint paper studies the mean cumulative number of recurrent events when an event such as death terminates further recurrence. Its contribution is statistical estimation and inference for a stated population quantity, rather than assuming that each observed event is an independent new person.[ref]
Ghosh’s 2010 single-author paper further examines recurrent-event analysis with dependent censoring, linking artificial-censoring methods with truncated-data structures and pairwise estimation. The point is to state and handle a particular observation mechanism, not to claim that every missing-data problem has the same correction.[ref]
An original observation example explains the motivation. If individuals with many events are more likely to stop being followed, an unadjusted mean of observed counts can look low because later events are hidden. If a terminal event is actually observed, later events are precluded instead. Recording why observation ends is therefore part of interpreting the data, alongside the count itself. These distinctions connect Ghosh’s documented work to survival analysis, counting processes and reproducible statistical definitions.
Debashis Ghosh
. Portrait linked from the named Colorado School of Public Health directory profile. Articles by others on the same topic
There are currently no matching articles.