Wireless Dataset Similarity: Measuring Distances in Supervised and Unsupervised Machine Learning

João Morais, Sadjad Alikhani, Akshay Malhotra, Shahab Hamidi-Rad, Ahmed Alkhateeb{joao, alikhani, alkhateeb}@asu.edu , {akshay.malhotra, shahab.hamidi-rad}@interdigital.com · 2026

Read Wireless Dataset Similarity: Measuring Distances in Supervised and Unsupervised Machine Learning by João Morais, Sadjad Alikhani, Akshay Malhotra, Shahab Hamidi-Rad, Ahmed Alkhateeb{joao, alikhani, alkhateeb}@asu.edu , {akshay.malhotra, shahab.hamidi-rad}@interdigital.com (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-01023.txt.

Description

João Morais and four colleagues ask how alike two wireless data sets really are, and why that matters. A model trained on one set of radio data often fails on another, so teams need a way to judge fit before they train. Their framework scores the gap between data sets in a way that is tied to the task and the model at hand.

The key test is simple: a small gap should mean a model trained on one set works well on the other. The team tries the idea on signal squeezing with autoencoders and on beam choice with neural networks. Their scores track real transfer results far better than older measures. That helps teams pick data, build fake data, and adapt models to new sites with less guesswork.

Catalog record

CREATORJoão Morais, Sadjad Alikhani, Akshay Malhotra, Shahab Hamidi-Rad, Ahmed Alkhateeb{joao, alikhani, alkhateeb}@asu.edu , {akshay.malhotra, shahab.hamidi-rad}@interdigital.com
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE73 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-01023

Browse the full library · About the Tower of Babel