his.sePublikationer
Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Optimering av datamängder med Machine learning: En studie om Machine learning och Internet of Things
Högskolan i Skövde, Institutionen för informationsteknologi.
2017 (Svenska)Självständigt arbete på grundnivå (kandidatexamen), 20 poäng / 30 hpStudentuppsats (Examensarbete)Alternativ titel
Optimizing dataflow with Machine learning : a study about Machine learning and Internet of Things (Engelska)
Abstract [en]

This report is about how an Internet of Things (IoT) optimization can be done with Machine learning (ML). The IoT- devices in this report are sensors in containers that read how full the containers are. The report contains a case from Sogeti. Were a client can use this optimization to get better routes for their garbage truck, with this solution the garbage trucks will only go to full containers and be able to skip empty or close to empty containers. This will result in less fuel costs and a better environment. This solution can be used for every industry that needs a route optimization. To do this there must first be understanding for what IoT is and what is possible to do with it then there need to be understanding about ML. The report cover these parts and tell how the method Design science (DS) is being used to produce this solution and some information about the method. This project also works agile with iterations under the implementation stage in DS. On the ML part there is an argumentation of a comparison of witch algorithm should be used. There are two candidates: Hill- Climbing and K-means Cluster. For this solution K-means cluster will be the one being used. K-means clustering is an unsupervised algorithm that doesn’t need practice data, it pairs data that are very similar and builds clusters. It will do this with full containers and build clusters with the ones that have similar coordinates so the full containers are close to each other. When this is done the clusters exports to a database and then there is a brief description on how its possible to make a map that makes a route between the containers in the cluster. 

Ort, förlag, år, upplaga, sidor
2017. , s. 37
Nyckelord [sv]
Machine learning, Internet of Things, cluster, Rutt optimering, rutt optimering med machine learning, ML, IoT
Nationell ämneskategori
Systemvetenskap, informationssystem och informatik
Identifikatorer
URN: urn:nbn:se:his:diva-13747OAI: oai:DiVA.org:his-13747DiVA, id: diva2:1111127
Externt samarbete
Sogeti
Ämne / kurs
Informationsteknologi
Utbildningsprogram
Systemvetenskap - inriktning Enterprise Information Management
Handledare
Examinatorer
Tillgänglig från: 2017-07-09 Skapad: 2017-06-17 Senast uppdaterad: 2018-01-13Bibliografiskt granskad

Open Access i DiVA

fulltext(1185 kB)186 nedladdningar
Filinformation
Filnamn FULLTEXT01.pdfFilstorlek 1185 kBChecksumma SHA-512
910324814397895970080ccd80a00be1fe52d5f0cc8bd712e144e148a5583746f312a817fc4f27618e10ea542ed6281159cf34f418de8f1f6049511ab8c9ce33
Typ fulltextMimetyp application/pdf

Av organisationen
Institutionen för informationsteknologi
Systemvetenskap, informationssystem och informatik

Sök vidare utanför DiVA

GoogleGoogle Scholar
Totalt: 186 nedladdningar
Antalet nedladdningar är summan av nedladdningar för alla fulltexter. Det kan inkludera t.ex tidigare versioner som nu inte längre är tillgängliga.

urn-nbn

Altmetricpoäng

urn-nbn
Totalt: 576 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf