ResBit: Residual Bit Vector for Categorical Values

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Fuchi, Masane, Zanashir, Amar, Minami, Hiroto, Takagi, Tomohiro
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917836649136128
author Fuchi, Masane
Zanashir, Amar
Minami, Hiroto
Takagi, Tomohiro
author_facet Fuchi, Masane
Zanashir, Amar
Minami, Hiroto
Takagi, Tomohiro
contents One-hot vectors, a common method for representing discrete/categorical data, in machine learning are widely used because of their simplicity and intuitiveness. However, one-hot vectors suffer from a linear increase in dimensionality, posing computational and memory challenges, especially when dealing with datasets containing numerous categories. In this paper, we focus on tabular data generation, and reveal the multinomial diffusion faces the mode collapse phenomenon when the cardinality is high. Moreover, due to the limitations of one-hot vectors, the training phase takes time longer in such a situation. To address these issues, we propose Residual Bit Vectors (ResBit), a technique for densely representing categorical data. ResBit is an extension of analog bits and overcomes limitations of analog bits when applied to tabular data generation. Our experiments demonstrate that ResBit not only accelerates training but also maintains performance when compared with the situations before applying ResBit. Furthermore, our results indicate that many existing methods struggle with high-cardinality data, underscoring the need for lower-dimensional representations, such as ResBit and latent vectors.
format Preprint
id arxiv_https___arxiv_org_abs_2309_17196
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle ResBit: Residual Bit Vector for Categorical Values
Fuchi, Masane
Zanashir, Amar
Minami, Hiroto
Takagi, Tomohiro
Machine Learning
One-hot vectors, a common method for representing discrete/categorical data, in machine learning are widely used because of their simplicity and intuitiveness. However, one-hot vectors suffer from a linear increase in dimensionality, posing computational and memory challenges, especially when dealing with datasets containing numerous categories. In this paper, we focus on tabular data generation, and reveal the multinomial diffusion faces the mode collapse phenomenon when the cardinality is high. Moreover, due to the limitations of one-hot vectors, the training phase takes time longer in such a situation. To address these issues, we propose Residual Bit Vectors (ResBit), a technique for densely representing categorical data. ResBit is an extension of analog bits and overcomes limitations of analog bits when applied to tabular data generation. Our experiments demonstrate that ResBit not only accelerates training but also maintains performance when compared with the situations before applying ResBit. Furthermore, our results indicate that many existing methods struggle with high-cardinality data, underscoring the need for lower-dimensional representations, such as ResBit and latent vectors.
title ResBit: Residual Bit Vector for Categorical Values
topic Machine Learning
url https://arxiv.org/abs/2309.17196