A Simple Baseline for Predicting Events with Auto-Regressive Tabular Transformers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Stein, Alex, Sharpe, Samuel, Bergman, Doron, Kumar, Senthil, Bruss, C. Bayan, Dickerson, John, Goldstein, Tom, Goldblum, Micah
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915001422315520
author Stein, Alex
Sharpe, Samuel
Bergman, Doron
Kumar, Senthil
Bruss, C. Bayan
Dickerson, John
Goldstein, Tom
Goldblum, Micah
author_facet Stein, Alex
Sharpe, Samuel
Bergman, Doron
Kumar, Senthil
Bruss, C. Bayan
Dickerson, John
Goldstein, Tom
Goldblum, Micah
contents Many real-world applications of tabular data involve using historic events to predict properties of new ones, for example whether a credit card transaction is fraudulent or what rating a customer will assign a product on a retail platform. Existing approaches to event prediction include costly, brittle, and application-dependent techniques such as time-aware positional embeddings, learned row and field encodings, and oversampling methods for addressing class imbalance. Moreover, these approaches often assume specific use-cases, for example that we know the labels of all historic events or that we only predict a pre-specified label and not the data's features themselves. In this work, we propose a simple but flexible baseline using standard autoregressive LLM-style transformers with elementary positional embeddings and a causal language modeling objective. Our baseline outperforms existing approaches across popular datasets and can be employed for various use-cases. We demonstrate that the same model can predict labels, impute missing values, or model event sequences.
format Preprint
id arxiv_https___arxiv_org_abs_2410_10648
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Simple Baseline for Predicting Events with Auto-Regressive Tabular Transformers
Stein, Alex
Sharpe, Samuel
Bergman, Doron
Kumar, Senthil
Bruss, C. Bayan
Dickerson, John
Goldstein, Tom
Goldblum, Micah
Machine Learning
Computational Engineering, Finance, and Science
Many real-world applications of tabular data involve using historic events to predict properties of new ones, for example whether a credit card transaction is fraudulent or what rating a customer will assign a product on a retail platform. Existing approaches to event prediction include costly, brittle, and application-dependent techniques such as time-aware positional embeddings, learned row and field encodings, and oversampling methods for addressing class imbalance. Moreover, these approaches often assume specific use-cases, for example that we know the labels of all historic events or that we only predict a pre-specified label and not the data's features themselves. In this work, we propose a simple but flexible baseline using standard autoregressive LLM-style transformers with elementary positional embeddings and a causal language modeling objective. Our baseline outperforms existing approaches across popular datasets and can be employed for various use-cases. We demonstrate that the same model can predict labels, impute missing values, or model event sequences.
title A Simple Baseline for Predicting Events with Auto-Regressive Tabular Transformers
topic Machine Learning
Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2410.10648