Does data splitting improve prediction?

DOI10.1007/S11222-014-9522-9zbMATH Open1342.62025arXiv1301.2983OpenAlexW3125270795MaRDI QIDQ2631345FDOQ2631345

Authors: Julian J. Faraway 

Publication date: 29 July 2016

Published in: Statistics and Computing (Search for Journal in Brave)

Abstract: Data splitting divides data into two parts. One part is reserved for model selection. In some applications, the second part is used for model validation but we use this part for estimating the parameters of the chosen model. We focus on the problem of constructing reliable predictive distributions for future observed values. We judge the predictive performance using log scoring. We compare the full data strategy with the data splitting strategy for prediction. We show how the full data score can be decomposed into model selection, parameter estimation and data reuse costs. Data splitting is preferred when data reuse costs are high. We investigate the relative performance of the strategies in four simulation scenarios. We introduce a hybrid estimator called SAFE that uses one part for model selection but both parts for estimation. We discuss the choice to use a split data analysis versus a full data analysis.

Full work available at URL: https://arxiv.org/abs/1301.2983

Recommendations

• Data partition methodology for validation of predictive models
• How much better is disaggregate data?
• Forecasting with many predictors: is boosting a viable alternative?
• Multi split conformal prediction
• The effect of splitting on random forests
• Does adding data always improve linear regression estimates?

zbMATH Keywords

predictioncross-validationmodel assessmentmodel uncertaintyscoringmodel validation

Mathematics Subject Classification ID

Point estimation (62F10) General considerations in statistical decision theory (62C05)

Cites Work

• Statistical Analysis of Financial Data in S-Plus
• The elements of statistical learning. Data mining, inference, and prediction
• Strictly Proper Scoring Rules, Prediction, and Estimation
• Confidence distribution, the frequentist distribution estimator of a parameter: a review
• Least squares after model selection in high-dimensional sparse models
• Present Position and Potential Developments: Some Personal Views: Statistical Theory: The Prequential Approach
• MODEL SELECTION AND INFERENCE: FACTS AND FICTION
• Title not available (Why is that?)
• Frequentist prediction intervals and predictive distributions
• Title not available (Why is that?)
• Split Samples and Design Sensitivity in Observational Studies
• Clinical prediction models. A practical approach to development, validation, and updating.
• The Analysis of Transformed Data
• Proper local scoring rules
• Title not available (Why is that?)
• A note on data-splitting for the evaluation of significance levels
• Cross-Validation of Regression Models

Cited In (2)

• How much better is disaggregate data?
• Statistical Inference Enables Bad Science; Statistical Thinking Enables Good Science

Uses Software

• R
• MASS (R)

This page was built for publication: Does data splitting improve prediction?

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2631345)