Mjerenje i analiza korelacija WiFi signala i broja ljudi pomoću strojnog učenja

master's thesis

• Title (english): Measurement and analysis of correlations between WiFi signal and a number of people via machine learning
• Author: Karlo Kumiša
• Mentor: Hrvoje Buljan (mentor)
• Committee member: Hrvoje Buljan (predsjednik povjerenstva)
• Committee member: Davor Horvatić (član povjerenstva)
• Committee member: Katarina Jeličić (član povjerenstva)
• Committee member: Goranka Bilalbegović (član povjerenstva)
• Committee member: Maja Planinić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Physics); Zagreb
• Defense date and country: 2022-09-28, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Physics

Abstract

U ovom radu istražuje se mogućnost predviđanja broja ljudi u danom prostoru kroz jakost Wi-Fi polja u određenim točkama prostora. Predviđanja ostvaruju algoritmi strojnog učenja iz podataka generiranih kroz simulaciju i eksperiment. Iako rezultati na eksperimentalnim podacima nisu zadovoljavajući, istraživanje na podacima iz simulacije daje obećavajuće rezultate s obzirom na to da je najveća točnost predviđanja oko 69%.

Abstract (english)

This paper investigates the possibility of predicting the number of people in a given space through the strength of Wi-Fi signal in certain points of space. Predictions are made by machine learning algorithms from data generated through simulation and experiment. Although the results of the experiment data are not satisfactory, research on simulation data give promising results considering that the highest prediction accuracy is around 69%.

• Keywords: Wi-Fi; nadzirano učenje; strojno učenje; senzori; stablo odlučivanja; k najbližih susjeda; neuralna mreža; konvolucijska neuralna mreža
• Keywords (english): Wi-Fi; supervised learning; machine learning; sensors; decision tree; k nearest neighbors; neural network; convolutional neural network
• Language: croatian
• URN:NBN: urn:nbn:hr:217:963152
• Study programme: Title: Physics; specializations in: Education; Course: Education; Study programme type: university; Study level: integrated undergraduate and graduate; Academic / professional title: magistar/magistra edukacije fizike (magistar/magistra edukacije fizike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access; Embargo expiration date: 2024-03-28
• Repository: Repository of the Faculty of Science
• Created on: 2022-10-20 09:41:41