Saved in:
Bibliographic Details
Main Authors: Zhen, Cheng, Aryal, Nischal, Termehchy, Arash, Aghasi, Alireza, Chabada, Amandeep Singh
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2402.17926
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910350089125888
author Zhen, Cheng
Aryal, Nischal
Termehchy, Arash
Aghasi, Alireza
Chabada, Amandeep Singh
author_facet Zhen, Cheng
Aryal, Nischal
Termehchy, Arash
Aghasi, Alireza
Chabada, Amandeep Singh
contents Real-world data is often incomplete and contains missing values. To train accurate models over real-world datasets, users need to spend a substantial amount of time and resources imputing and finding proper values for missing data items. In this paper, we demonstrate that it is possible to learn accurate models directly from data with missing values for certain training data and target models. We propose a unified approach for checking the necessity of data imputation to learn accurate models across various widely-used machine learning paradigms. We build efficient algorithms with theoretical guarantees to check this necessity and return accurate models in cases where imputation is unnecessary. Our extensive experiments indicate that our proposed algorithms significantly reduce the amount of time and effort needed for data imputation without imposing considerable computational overhead.
format Preprint
id arxiv_https___arxiv_org_abs_2402_17926
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Certain and Approximately Certain Models for Statistical Learning
Zhen, Cheng
Aryal, Nischal
Termehchy, Arash
Aghasi, Alireza
Chabada, Amandeep Singh
Machine Learning
Databases
Real-world data is often incomplete and contains missing values. To train accurate models over real-world datasets, users need to spend a substantial amount of time and resources imputing and finding proper values for missing data items. In this paper, we demonstrate that it is possible to learn accurate models directly from data with missing values for certain training data and target models. We propose a unified approach for checking the necessity of data imputation to learn accurate models across various widely-used machine learning paradigms. We build efficient algorithms with theoretical guarantees to check this necessity and return accurate models in cases where imputation is unnecessary. Our extensive experiments indicate that our proposed algorithms significantly reduce the amount of time and effort needed for data imputation without imposing considerable computational overhead.
title Certain and Approximately Certain Models for Statistical Learning
topic Machine Learning
Databases
url https://arxiv.org/abs/2402.17926