Electricity Companies as Healthcare Service

No Thumbnail Available

Files

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Bachelor's thesis
Electronic archive copy is available locally at the Harald Herlin Learning Centre. The staff of Aalto University has access to the electronic bachelor's theses by logging into Aaltodoc with their personal Aalto user ID. Read more about the availability of the bachelor's theses.
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2024-04-26

Department

Major/Subject

Data Science

Mcode

SCI3095

Degree programme

Aalto Bachelor’s Programme in Science and Technology

Language

en

Pages

41+4

Series

Abstract

The potential of applying machine learning and artificial intelligence in healthcare services has been vigorously studied in recent years. These technologies possess strong predictive power, which is highly beneficial for preventive measures, including early disease detection or interactive education of healthy habits. However, privacy concerns and the need for extensive dataset hinder the widespread adoption of these applications. One solution to such an issue is to utilise the readily available electricity data as a non-intrusive monitoring method for detecting certain harmful behaviours. The link between human behaviours and electricity consumption has been explored with various approaches, such as studying behav-iours as drivers towards energy usage or non-intrusive appliance load monitoring. One of these approaches is occupancy detection, which aims to infer occupancy status from power load. This thesis reviews relevant literature and conducts an experiment that focuses on modelling occupancy from electricity consumption. The exper-iment uses the publicly available Electricity Consumption & Occupancy (ECO) dataset and examines popular modelling choices for occupancy detection: support vector machine (SVM), k-nearest neighbours (KNN), multilayer perceptron (MLP), random forest (RF), and gradient boosting (GB). Amongst these models, SVM, RF, and GB emerge as the best-performing models that consistently achieve high accuracy scores above 0.7 on validation dataset, as well as acceptable ROC AUC and specificity scores. These results are highly promising and encourages further re-search into the relationship between human behaviours and energy usage.

Description

Supervisor

Korpi-Lagg, Maarit

Thesis advisor

Jung, Alexander

Keywords

occupancy detection, smart meters, electricity load curves, machine learning, support vector machine, K-nearest neighbour

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202405283856

Collections

[kand] Perustieteiden korkeakoulu / SCI