Ponedeljkov seminar računalništva in informatike - Arhiv
| 2024 | 2023 | 2022 | 2021 | 2020 | 2019 | 2018 | 2017 | 2016 | 2015 | 2014 | 2013 | 2012 |
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 |
sreda, 26. julij 2023 Gregor GRBEC: Ranking Football Players Using Multilevel Modeling
V ponedeljek, 31. julija 2023, bo ob 16.00 uri izvedeno
predavanje v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE
Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM.
ČAS/PROSTOR: 31. julij 2023 ob 16.00 na daljavo prek ZOOM-a
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
----------------------------------------
PREDAVATELJ: Gregor GRBEC
----------------------------------------
Gregor Grbec is a student who started his studies in Bioinformatics at UP FAMNIT. After graduation in 2020, he decided to pursue a master's degree in Data Science at the same faculty. His lifelong interest in football inspired him to do research on who is the best football player – one of the most frequently asked questions in football.
------------------------------------------------------------------------------------
NASLOV: Ranking Football Players Using Multilevel Modeling
------------------------------------------------------------------------------------
POVZETEK:
While football is a team sport, the question of who is the best football player is one of the most frequently asked in the footballing world. Attackers like Pele, Maradona, Cristiano Ronaldo and Lionel Messi are often mentioned as the best of all time, while players of other positions are mentioned significantly less. We believe that players of all positions are important, which is why we wanted to provide a data-backed answer to this question using advanced mathematical models that can incorporate players of all positions on a single list. Our two main objectives were to extract an individual's performance from the team's performance and to eliminate any bias that comes from the player's position. We used multilevel modeling which allowed us to control for player's team strength with data collected for all players of the top 20 clubs from the top 5 leagues from this century. The results were surprising, as Messi and Ronaldo, often believed to be two of the best of all time, ranked lower than expected.
Seminar bo potekal v angleškem jeziku, tokrat na daljavo prek Zoom-a
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
Vabljeni!
ponedeljek, 24. julij 2023 Marina PALDAUF: Decentralised Solutions for Preserving Privacy in Group Recommender Systems
V ponedeljek, 24. julija 2023, bo ob 17.00 uri izvedeno
predavanje v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE
Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM.
ČAS/PROSTOR: 24. julij 2023 ob 17.00 na daljavo prek ZOOM-a
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
------------------------------------------------
PREDAVATELJICA: Marina PALDAUF
------------------------------------------------
Marina Paldauf is a PhD student in Computer Science in the fields of recommender systems, machine learning and distributed systems and a teaching assistant at the University of Primorska, Faculty of Mathematics, Natural Sciences and Information Technologies. She is also a researcher at the HICUP Lab and a Department of Information Sciences and Technologies member.
-------------------------------------------------------------------------------------------------------------------------
NASLOV: Decentralised Solutions for Preserving Privacy in Group Recommender Systems
-------------------------------------------------------------------------------------------------------------------------
POVZETEK:
Group Recommender Systems (GRS) combine large amounts of data from various user behaviour signals (likes, views, purchases) and contextual information to provide groups of users with accurate suggestions (e.g. rating prediction, rankings). To handle those large amounts of data, GRS can be extended to use distributed processing and storage solutions (e.g. MapReduce-like algorithms and NoSQL databases). As such, privacy has always been a core issue since most recommendation algorithms rely on user behaviour signals and contextual information that may contain sensitive information. However, existing work in this domain mostly distributes data processing tasks without addressing privacy, and the solutions that address privacy for GRS (e.g. k-anonymisation and local differential privacy) remain centralised. In this paper, we identify and analyse privacy concerns in GRS and provide guidelines on how decentralised techniques can be used to address them.
Seminar bo potekal v angleškem jeziku, tokrat na daljavo prek Zoom-a s pričetkom ob 17:00 uri
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
Vabljeni!
petek, 14. julij 2023 Aleksandar AVDALOVIĆ: Translation of random forests to loop-free imperative programs for the purpose of formal verification
V ponedeljek, 17. julija 2023, bo ob 17.00 uri izvedeno
predavanje v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE
Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM.
ČAS/PROSTOR: 17. julij 2023 ob 17.00 na daljavo prek ZOOM-a
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
---------------------------------------------------
PREDAVATELJ: Aleksandar AVDALOVIĆ
---------------------------------------------------
Aleksandar Avdalović is a second-year student of the master's program Data Science at UP FAMNIT. He will be presenting his Master’s thesis topic that he is completing under the supervision of Prof. Mário Alberto Zenha-Rela and Prof. Raul Barbosa from University of Coimbra and co-supervision of Prof. Marko Tkalčič from UP FAMNIT.
------------------------------------------------------------------------------------------------------------------------------------------------------
NASLOV: Translation of random forests to loop-free imperative programs for the purpose of formal verification
------------------------------------------------------------------------------------------------------------------------------------------------------
POVZETEK:
Throughout the years of society's digitalisation and transformation, the ethics of AI/ML-dependent critical systems have been continually questioned. In the use of such systems, accuracy may not be a valid metric because machine errors can lead to devastating consequences, including the loss of human life. This thesis takes a narrow approach to the aforementioned problem, namely the formal verification of the incredibly effective and widespread supervised learning method of today, random forest. Using the first-order logic statements and algorithm developed in this study, we intend to determine if the model has acquired factually-grounded prior knowledge. In other words, we wish to confirm that the model is reliable in light of prior knowledge. The translation of the random forest algorithm into a loop-free imperative program for verification purposes provides a technical challenge for this task. The core of the thesis is the algorithm provided in response to the research question of whether an algorithm for described translation exists. We shall construct the algorithm and ensure its correctness by rigorously testing it with various datasets, including those later used for verification and artificially generated datasets with different shapes and numbers of classes. Additionally, we will evaluate its efficiency and demonstrate its application on three datasets (one dataset from Cardiology and two datasets from Rheumatology). Another research problem will be practical, examining if the model has acquired particular prior knowledge in the fields of Cardiology and Rheumatology given the data. Results demonstrate that such an algorithm can be created and is highly efficient and that it can be applied realistically in the health sector to validate ML models based on the random forest algorithm.
Seminar bo potekal v angleškem jeziku, tokrat na daljavo prek Zoom-a
(https://upr-si.zoom.us/j/297328207?pwd=S3Zpdk1VR3pjckNtWkQwKzlvcDR5UT09)
Vabljeni!
ponedeljek, 10. julij 2023 Nuwan ATTYGALLE: Breaking Barriers in Gesture Recognition: Enhancing Accuracy through Material-Aware Training
V ponedeljek, 10. julija 2023, bo ob 16.00 uri izvedeno
predavanje v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE
Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM.
ČAS/PROSTOR: 10. julij 2023 ob 16.00 v FAMNIT-VP2.
----------------------------------------------
PREDAVATELJ: Nuwan ATTYGALLE
----------------------------------------------
Nuwan Attygalle is a PhD student at UP FAMNIT with a research focus at the intersection of signal processing, HCI, and machine learning. His interests lie in exploring innovative applications and solutions in these fields. Nuwan has also served as a visiting researcher at the COIN lab at the University of Luxembourg, further expanding his knowledge and expertise in his research areas.
----------------------------------------------------------------------------------------------------------------------------------------------
NASLOV: Breaking Barriers in Gesture Recognition: Enhancing Accuracy through Material-Aware Training
----------------------------------------------------------------------------------------------------------------------------------------------
POVZETEK:
Human-computer interaction is undergoing a transformative shift, moving away from traditional interfaces (screen, keyboard and mouse) and towards more intuitive and natural interaction methods. Gesture recognition plays a crucial role in this paradigm shift, enabling participants to interact with digital devices using natural hand movements and gestures. Leveraging millimetre-wave (mmWave) radar signals that penetrate materials and are immune to environmental conditions, gesture recognition can be achieved even when devices are concealed or embedded within objects. However, the impact of everyday materials on gesture recognition accuracy remains inadequately investigated. This paper investigates whether training a machine learning model with gestures sensed through materials improves overall classification accuracy. We also explore to what extent this improvement occurs and how to achieve comparable accuracy with a model trained using fewer materials. Our findings shed light on the influence of materials on gesture recognition and provide insights into optimizing training processes for enhanced performance in real-world applications.
Seminar bo potekal v angleškem jeziku v predavalnici FAMNIT-VP2.
Vabljeni!
