Saved in:
Bibliographic Details
Main Authors: Castellanos, Nathalia, Desai, Dhruv, Frank, Sebastian, Pasquali, Stefano, Mehta, Dhagash
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.10340
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911995663482880
author Castellanos, Nathalia
Desai, Dhruv
Frank, Sebastian
Pasquali, Stefano
Mehta, Dhagash
author_facet Castellanos, Nathalia
Desai, Dhruv
Frank, Sebastian
Pasquali, Stefano
Mehta, Dhagash
contents Peer analysis is a critical component of investment management, often relying on expert-provided categorization systems. These systems' consistency is questioned when they do not align with cohorts from unsupervised clustering algorithms optimized for various metrics. We investigate whether unsupervised clustering can reproduce ground truth classes in a labeled dataset, showing that success depends on feature selection and the chosen distance metric. Using toy datasets and fund categorization as real-world examples we demonstrate that accurately reproducing ground truth classes is challenging. We also highlight the limitations of standard clustering evaluation metrics in identifying the optimal number of clusters relative to the ground truth classes. We then show that if appropriate features are available in the dataset, and a proper distance metric is known (e.g., using a supervised Random Forest-based distance metric learning method), then an unsupervised clustering can indeed reproduce the ground truth classes as distinct clusters.
format Preprint
id arxiv_https___arxiv_org_abs_2408_10340
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Can an unsupervised clustering algorithm reproduce a categorization system?
Castellanos, Nathalia
Desai, Dhruv
Frank, Sebastian
Pasquali, Stefano
Mehta, Dhagash
Machine Learning
Statistical Finance
Applications
Peer analysis is a critical component of investment management, often relying on expert-provided categorization systems. These systems' consistency is questioned when they do not align with cohorts from unsupervised clustering algorithms optimized for various metrics. We investigate whether unsupervised clustering can reproduce ground truth classes in a labeled dataset, showing that success depends on feature selection and the chosen distance metric. Using toy datasets and fund categorization as real-world examples we demonstrate that accurately reproducing ground truth classes is challenging. We also highlight the limitations of standard clustering evaluation metrics in identifying the optimal number of clusters relative to the ground truth classes. We then show that if appropriate features are available in the dataset, and a proper distance metric is known (e.g., using a supervised Random Forest-based distance metric learning method), then an unsupervised clustering can indeed reproduce the ground truth classes as distinct clusters.
title Can an unsupervised clustering algorithm reproduce a categorization system?
topic Machine Learning
Statistical Finance
Applications
url https://arxiv.org/abs/2408.10340