Please use this identifier to cite or link to this item: http://dr.iiserpune.ac.in:8080/xmlui/handle/123456789/7818
Title: Proved Algorithms in Geometric Group Theory
Authors: Gadgil, Siddhartha
TADIPATRI, ANAND RAO
Dept. of Mathematics
20181122
Keywords: Research Subject Categories::MATHEMATICS
Issue Date: May-2023
Citation: 97
Abstract: In the recent decades, computers have opened up novel ways of doing mathematics. The speed and raw power of computers has been used to perform searches over previously untractable spaces of objects and proofs, and complementarily, the ability of computers be systematic and methodical has been utilised to formalise and automatically verify complicated proofs. Although writing programs and constructing proofs seem like different activities, in type-theoretic foundations of mathematics these can both be seen as instances of constructing a term of a type. Proof assistants based on dependent type theory leverage this correspondence to serve both as theorem provers as well as functional programming languages. As programming and proving can be done within the same framework of dependent type theory, computers searches can in principle be augmented with proofs to give "proved algorithms" that have been proved to be correct and returns results together with formal proofs when run. Geometric group theory is a relatively new area that is at the intersection of algebra and topology, and connected to several others. Over the last few years, there have been some significant results in the area achieved with the help of computer searches. This thesis is centered on the theme of “Proved Algorithms in Geometric Group Theory”, using the Lean4 theorem prover and programming language to formally prove results in geometric group theory involving a mix of proofs and computations.
URI: http://dr.iiserpune.ac.in:8080/xmlui/handle/123456789/7818
Appears in Collections:MS THESES

Files in This Item:
File Description SizeFormat 
20181122_ANAND_TADIPATRI_MS_Thesis.pdfMS Thesis986.9 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.