Please use this identifier to cite or link to this item:
https://elib.bsu.by/handle/123456789/233372| Title: | Influence of the training set on the prediction stability in estimation of acute pancreatitis severity |
| Authors: | Mangalova, E. Chubarova, O. Melekh, D. |
| Keywords: | ЭБ БГУ::ЕСТЕСТВЕННЫЕ И ТОЧНЫЕ НАУКИ::Математика ЭБ БГУ::ЕСТЕСТВЕННЫЕ И ТОЧНЫЕ НАУКИ::Кибернетика |
| Issue Date: | 2019 |
| Publisher: | Minsk : BSU |
| Citation: | Computer Data Analysis and Modeling: Stochastics and Data Science : Proc. of the Twelfth Intern. Conf., Minsk, Sept. 18-22, 2019. – Minsk : BSU, 2019. – P. 232-236. |
| Abstract: | The small sample size problem is often encountered in data analysis, especially for medical applications. It leads to unstable predictions when including or excluding several observations could change prediction significantly. Prediction stability visualization and measure were proposed and applied to estimation of acute pancreatitis severity. A simulation experiments were carried out to study the stability of ridge-regression, SVM, random forest trained with various subsets |
| URI: | http://elib.bsu.by/handle/123456789/233372 |
| ISBN: | 978-985-566-811-5 |
| Appears in Collections: | 2019. Computer Data Analysis and Modeling : Stochastics and Data Science |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| 232-236.pdf | 446,57 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.