Abstract	Cryptanalysis exploits weaknesses in ciphers to recover hidden information; however, traditional techniques become increasingly inefficient as ciphers become more complex. This paper investigates supervised machine learning as a tool for classical cryptanalysis. Shift, Affine, and Vigen`ere ciphers were implemented and applied to chunks of text to generate training data. From each ciphertext, numerical features such as letter frequencies, entropy, and index of coincidence were developed and used to train models for key recovery. Across all three ciphers, the machine learning models achieved higher key recovery accuracy and efficiency than classical manual methods, with particularly strong improvements for the Vigen`ere cipher. These results demonstrate that machine learning can effectively support the cryptanalysis of classical ciphers, suggesting promising applications to more advanced ciphers.
Alternative title	University of Wisconsin-La Crosse Undergraduate Research & Creativity Laureate Program
Author	Klawitter, Charlie
Contributor	University of Wisconsin-La Crosse Undergraduate Research & Creativity Laureate Program
Owner	University of Wisconsin-La Crosse
Sponsor	Vidden, Chad
Type of resource	text
Genre	journal
Genre authority	marcgt
Publisher name	University of Wisconsin-La Crosse
Place of publication	La Crosse, Wisconsin
Date published	2025
Date captured	2026-03-09
Language	eng
Width	8.5
Height	10.99
Subject topic	University of Wisconsin-La Crosse -- Students -- Research -- Periodicals Authority: LCSH
Subject topic	College students -- Research -- United States -- Periodicals Authority: LCSH
Subject topic	Journals Authority: LCSH
Subject topic	Mathematics Authority: LCSH
Source note	Published as part of the University of Wisconsin-La Crosse Undergraduate Research & Creativity Laureate Program, Fall 2025.
Use and reproduction restrictions	This material may be protected by copyright law (e.g., Title 17, US Code). For more information about the University of Wisconsin-La Crosse Murphy Library's copyright, fair-use, and permissions policies, please see https://digitalcollections.uwlax.edu/.
Collection	UWL Undergraduate Research & Creativity Collection / UWL Undergraduate Research & Creativity Laureate Program
ID	da1b46a9-5f42-4062-a8cc-bc5486f51ee7/wlacu000/00000013/00000799
Doi	13-00799