Big data algorithm optimization

Examensarbete för masterexamen

Använd denna länk för att citera eller länka till detta dokument: https://hdl.handle.net/20.500.12380/184652
Ladda ner:
Fil Beskrivning StorlekFormat 
184652.pdfFulltext1.07 MBAdobe PDFThumbnail
Visa
Bibliografiska detaljer
FältVärde
Typ: Examensarbete för masterexamen
Master Thesis
Titel: Big data algorithm optimization
Författare: Karlsson, Kasper
Lans, Tobias
Sammanfattning: When sales representatives and customers negotiate, it must be confirmed that the final deals will render a high enough profit for the selling company. Large companies have different methods of doing this, one of which is to run sales simulations. Such simulation systems often need to perform complex calculations over large amounts of data, which in turn requires efficient models and algorithms. This project intends to evaluate whether it is possible to optimize and extend an existing sales system called PCT, which is currently suffering from unacceptably high running times in its simulation process. This is done through analysis of the current implementation, followed by optimization of its models and development of efficient algorithms. The performance of these optimized and extended models are compared to the existing one in order to evaluate their improvement. The conclusion of this project is that the simulation process in PCT can indeed be optimized and extended. The optimized models serve as a proof of concept, which shows that results identical to the original system's can be calculated within < 1% of the original running time for the largest customers.
Nyckelord: Data- och informationsvetenskap;Computer and Information Science
Utgivningsdatum: 2013
Utgivare: Chalmers tekniska högskola / Institutionen för data- och informationsteknik (Chalmers)
Chalmers University of Technology / Department of Computer Science and Engineering (Chalmers)
URI: https://hdl.handle.net/20.500.12380/184652
Samling:Examensarbeten för masterexamen // Master Theses



Materialet i Chalmers öppna arkiv är upphovsrättsligt skyddat och får ej användas i kommersiellt syfte!