Aristotle University Of Thessaloniki (AUTH)
Human-centered machine learning for mobile sensing data
Abstract
dc:descriptionThe present thesis delves into the convergence of mobile and wearable technologies, machine learning algorithms, and human-centered concerns within the Ubiquitous Computing domain. Ubiquitous devices have found numerous applications in recent years, including healthcare monitoring, workplace safety, quality control, remote assistance or training, and fatigue monitoring. Yet, the enthusiastic adoption of machine learning in such technologies may inadvertently perpetuate biases and ethical dilemmas. Through original research spanning user interaction, data collection, and algorithmic equity, this thesis proposes solutions to mitigate such “alignment” challenges. Among others, it introduces a framework for designing and evaluating technologies for behavior change interventions on diverse populations, open datasets and benchmarks for facilitating future interdisciplinary research, and methodologies for fairness-aware computing. By emphasizing the integration of technical advancements with human-centered values, this work advocates for a holistic approach to mobile and wearable computing, highlighting the importance of aligning machine learning systems with societal needs and expectations.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Yfantidou, Sofia
- Υφαντίδου, Σοφία
Subjects
dc:subject × 20- Μηχανική μάθηση
- Δεδομένα αισθητήρων
- Έξυπνες κινητές συσκευές
- Διεπαφή ανθρώπου-υπολογιστή
- Ανοιχτά δεδομένα
- Αμεροληψία
- Machine learning
- Sensing applications
- Mobile devices
- Wearable devices
- Human-computer interaction
- Open data
- ΑΙ fairness
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Τεχνητή νοημοσύνη
- Διάδραση ανθρώπου-υπολογιστή
- Computer and Information Sciences
- Natural Sciences
- Artificial Intelligence
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/56178