Saved in:
Bibliographic Details
Main Authors: Wang, Di, Shi, Junzhi, Wang, Pingping, Zhuang, Shuo, Li, Hongyue
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2301.04378
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910305190150144
author Wang, Di
Shi, Junzhi
Wang, Pingping
Zhuang, Shuo
Li, Hongyue
author_facet Wang, Di
Shi, Junzhi
Wang, Pingping
Zhuang, Shuo
Li, Hongyue
contents We propose a learning framework for calibrating predictive models to make loss-controlling prediction for exchangeable data, which extends our recently proposed conformal loss-controlling prediction for more general cases. By comparison, the predictors built by the proposed loss-controlling approach are not limited to set predictors, and the loss function can be any measurable function without the monotone assumption. To control the loss values in an efficient way, we introduce transformations preserving exchangeability to prove finite-sample controlling guarantee when the test label is obtained, and then develop an approximation approach to construct predictors. The transformations can be built on any predefined function, which include using optimization algorithms for parameter searching. This approach is a natural extension of conformal loss-controlling prediction, since it can be reduced to the latter when the set predictors have the nesting property and the loss functions are monotone. Our proposed method is applied to selective regression and high-impact weather forecasting problems, which demonstrates its effectiveness for general loss-controlling prediction.
format Preprint
id arxiv_https___arxiv_org_abs_2301_04378
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Loss-Controlling Calibration for Predictive Models
Wang, Di
Shi, Junzhi
Wang, Pingping
Zhuang, Shuo
Li, Hongyue
Machine Learning
We propose a learning framework for calibrating predictive models to make loss-controlling prediction for exchangeable data, which extends our recently proposed conformal loss-controlling prediction for more general cases. By comparison, the predictors built by the proposed loss-controlling approach are not limited to set predictors, and the loss function can be any measurable function without the monotone assumption. To control the loss values in an efficient way, we introduce transformations preserving exchangeability to prove finite-sample controlling guarantee when the test label is obtained, and then develop an approximation approach to construct predictors. The transformations can be built on any predefined function, which include using optimization algorithms for parameter searching. This approach is a natural extension of conformal loss-controlling prediction, since it can be reduced to the latter when the set predictors have the nesting property and the loss functions are monotone. Our proposed method is applied to selective regression and high-impact weather forecasting problems, which demonstrates its effectiveness for general loss-controlling prediction.
title Loss-Controlling Calibration for Predictive Models
topic Machine Learning
url https://arxiv.org/abs/2301.04378