Show simple item record

dc.contributor.advisorAndriamanalimanana, Bruno; Committee Chair
dc.contributor.advisorChiang, Chen-Fu; Thesis Committee
dc.contributor.advisorNovillo, Jorge; Thesis Committee
dc.contributor.authorYan, Jianzhi
dc.date.accessioned2021-03-09T16:45:49Z
dc.date.available2021-03-09T16:45:49Z
dc.date.issued2020-05
dc.identifier.urihttp://hdl.handle.net/20.500.12648/1655
dc.description.abstractThis project would give a comprehensive picture of non-convex optimization for deep learning, explain in details about Long Short-Term Memory (LSTM) and RMSProp. We start by illustrating the internal mechanisms of LSTM, like the network structure and backpropagation through time (BPTT). Then introducing RMSProp optimization, some relevant mathematical theorems and proofs in those sections, which give a clear picture of how RMSProp algorithm is helpful to escape the saddle point. After all the above, we apply it with LSTM with RMSProp for the experiment; the result would present the efficiency and accuracy, especially how our method beat traditional strategy in non-convex optimization.en_US
dc.language.isoen_USen_US
dc.subjectNonconvex programmingen_US
dc.subjectLong Short-Term Memory (LSTM)en_US
dc.subjectBack propagation (Artificial intelligence)en_US
dc.subjectRMSProp optimizationen_US
dc.titleNon-Convex Optimization: RMSProp Based Optimization for Long Short-Term Memory Networken_US
dc.typeThesisen_US
dc.description.versionNAen_US
refterms.dateFOA2021-03-09T16:45:50Z
dc.description.institutionSUNY Polytechnic Instituteen_US
dc.description.departmentDepartment of Computer Science and Software Engineeringen_US
dc.description.degreelevelMSen_US


Files in this item

Thumbnail
Name:
Jianzhi Yan FINAL document.pdf
Size:
1.624Mb
Format:
PDF
Description:
Final Thesis Submission
Thumbnail
Name:
J Yan signed dec.pdf
Size:
17.45Kb
Format:
PDF
Description:
Declaration
Thumbnail
Name:
J Yan signed adv sheet.pdf
Size:
22.15Kb
Format:
PDF
Description:
Advisor Sheet

This item appears in the following Collection(s)

Show simple item record