[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126244-en":3,"doc-seo-126244-105":31,"detail-sidebar-cat-0-en-105":93},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},126244,2336475104042,"Skyler","https://ap-avatar.wpscdn.com/avatar/22000c4c32af1715be0?x-image-process=image/resize,m_fixed,w_180,h_180&k=1786537525561427321",8,"Research & Report","An Exploration of Machine Learning Techniques for Securing Data Communication - Project Thesis","Securing communication addresses the long-standing problem of protecting transmitted data, commonly studied through cryptography. Building on recent machine learning advances, this project explores two ML-based cryptographic approaches: noisy training, which adds randomly generated noise, and adversarial neural cryptography, implemented via generative adversarial networks. The work explains how each method functions, details practical implementation, and evaluates performance trade-offs. In this study, noisy training runs roughly 1,000x faster but produces about 15% incorrect inferences, while the GAN approach enables reliable data recovery with high fidelity, informing appropriate method choice for specific applications.","CALIFORNIA STATE UNIVERSITY SAN MARCOS  \nPROJECT SIGNATURE PAGE  \nPROJECT SUBMITTED IN PARTIAL FULFILLMENT  \nOF THE REQUIREMENTS FOR THE DEGREE  \nMASTER OF SCIENCE  \nIN  \nCOMPUTER SCIENCE  \nPROJECT TITLE: An Exploration of Machine Learning Techniques for Securing Data Communication.  \nAUTHOR: James Erickson  \nDATE OF SUCCESSFUL DEFENSE: 03/29/23  \nTHE PROJECT HAS BEEN ACCEPTED BY THE PROJECT COMMITTEE IN  \nPARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN COMPUTER SCIENCE.  \nDr. Ali Ahmadinia  \nPROJECT COMMITTEE CHAIR  \nDr. Sreedevi Gutta  \nPROJECT COMMITTEE MEMBER  \nSIGNATURE  \nDATE  \nAN EXPLORATION OF MACHINE LEARNING TECHNIQUES FOR SECURING DATA COMMUNICATION  \nby  \nJames Erickson  \nA project  \nsubmitted in partial fulfillment  \nof the requirements for the degree of  \nMaster of Science in Computer Science  \nCalifornia State University San Marcos  \nFebruary 2023  \nABSTRACT  \nSecuring communication is a problem that has been widely studied throughout history. Indeed, an entire branch of mathematics, called cryptography, has been dedicated to solving this problem in various ways. Given this, and the recent advent of various Machine Learning techniques, it should be no surprise that efforts have been made to use ML techniques for cryptographic purposes. These efforts have resulted in multiple new techniques, two of which will be explored in this paper. The first involves adding randomly generated noise and is called noisy training. The second technique is known as adversarial neural cryptography, and involves generative adversarial networks. In looking at these methods, we describe in detail how they work, as well as discussing details of their implementation. We then compare them and discuss their advantages and disadvantages. With the implementations used in this paper, noisy training techniques are roughly 1,000x faster than adversarial neural cryptography. However, even with our best efforts in training, noisy training will make incorrect inferences about the sent data roughly 15% of the time. This is in contrast to adversarial neural cryptography, where our data sent via can be reliably recovered with a high degree of fidelity. Knowing about the differences between these approaches can be helpful in deciding what approach will be appropriate for a particular application.  \nTABLE OF CONTENTS  \nAN EXPLORATION OF MACHINE LEARNING TECHNIQUES FOR SECURING DATA COMMUNICATION ......... ii  \nABSTRACT............................................................................................................................................ 3  \n1. Introduction................................................................................................................................. 7  \n2. Related Work............................................................................................................................... 7  \n3. Average Common Sub-Matrix (ACSM) ........................................................................................ 10  \n3.1 Motivation for using ACSM................................................................................................. 10  \n3.2. Formal Definition of ACSM ................................................................................................. 11  \n4. Adversarial Neural Cryptography using GANs ............................................................................. 17  \n5. Results from the GAN-based approach....................................................................................... 21  \n6. The Noisy Training approach...................................................................................................... 22  \n7. Run Times on the Raspberry Pi versus the Training Machine ...................................................... 29  \n8. Comparison between the GAN-based approach and Noisy Training ........................................... 30  \n9. Future Work ................................................","cbCais1gc2I4UO1z","https://ap.wps.com/l/cbCais1gc2I4UO1z","pdf",902400,5,1,32,"English","en",105,"# Introduction\n# Related Work\n# Average Common Sub-Matrix (ACSM)\n## Motivation for using ACSM\n## Formal Definition of ACSM\n# Adversarial Neural Cryptography using GANs\n# Results from the GAN-based approach\n# The Noisy Training approach\n# Run Times on the Raspberry Pi versus the Training Machine\n# Comparison between the GAN-based approach and Noisy Training\n# Future Work\n# Conclusion\n# References","[{\"question\":\"What two machine learning techniques are evaluated for securing data communication?\",\"answer\":\"The project evaluates noisy training and adversarial neural cryptography using generative adversarial networks (GANs). Noisy training introduces randomly generated noise, while the GAN approach uses adversarial learning to secure communication.\"},{\"question\":\"How does the project assess performance differences between the two approaches?\",\"answer\":\"The document compares how the approaches function and their practical implementation details, then evaluates runtime and inference/recovery quality. It reports large runtime differences and differences in correctness and recoverability under training conditions.\"},{\"question\":\"What trade-offs are reported for noisy training versus the GAN-based adversarial approach?\",\"answer\":\"Noisy training is reported to be roughly 1,000x faster, but it yields incorrect inferences about sent data around 15% of the time. The GAN-based method is reported to recover data with high fidelity and more reliable reconstruction in the study’s setup.\"}]","An Exploration of Machine Learning Techniques for Securing Data Communication - Project Thesis | PDF",1785904023,81,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":88,"head_meta":90,"extra_data":92,"updated_unix":29},"an-exploration-of-machine-learning-techniques-for-securing-data-communication-project-thesis","",{"@graph":37,"@context":87},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/an-exploration-of-machine-learning-techniques-for-securing-data-communication-project-thesis/126244/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-08-24","2026-08-05",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83],{"name":74,"@type":75,"acceptedAnswer":76},"What two machine learning techniques are evaluated for securing data communication?","Question",{"text":77,"@type":78},"The project evaluates noisy training and adversarial neural cryptography using generative adversarial networks (GANs). Noisy training introduces randomly generated noise, while the GAN approach uses adversarial learning to secure communication.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"How does the project assess performance differences between the two approaches?",{"text":82,"@type":78},"The document compares how the approaches function and their practical implementation details, then evaluates runtime and inference/recovery quality. It reports large runtime differences and differences in correctness and recoverability under training conditions.",{"name":84,"@type":75,"acceptedAnswer":85},"What trade-offs are reported for noisy training versus the GAN-based adversarial approach?",{"text":86,"@type":78},"Noisy training is reported to be roughly 1,000x faster, but it yields incorrect inferences about sent data around 15% of the time. The GAN-based method is reported to recover data with high fidelity and more reliable reconstruction in the study’s setup.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,103,107,111,116,121,124,129,132,136],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Exam",70,"exam",{"id":20,"doc_module":4,"doc_module_name":47,"category_name":108,"show_sort_weight":109,"slug":110},"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":47,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":47,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":47,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":47,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":47,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":47,"category_name":138,"show_sort_weight":20,"slug":139},19,"General","general"]