Ponedeljkov seminar računalništva in informatike - Arhiv
| 2025 | 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 |
ponedeljek, 19. maj 2025 Vanja MILESKI: Deep learning for time series data using Inception and ResNet paradigms
V ponedeljek, 19. maja 2025, 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: 19. maj 2025 ob 16.00 v FAMNIT-VP2.
----------------------------------------
PREDAVATELJ: Vanja MILESKI
----------------------------------------
After graduating from the Faculty of Computer and Information Science at the University of Ljubljana in 2015, Vanja Mileski started working at the Jožef Stefan Institute (JSI). He was a Master's student at the International Postgraduate School Jožef Stefan and a student researcher at the JSI. After finishing his Master's studies, he applied his knowledge of data mining in the private sector as a Data Scientist in the retail, telecommunications, banking, stock market and insurance sectors. His current research interests include Time-Series classification, Deep Learning, ResNet and Inception architectures as well as LLMs.
-----------------------------------------------------------------------------------------------------------------
NASLOV: Deep learning for time series data using Inception and ResNet paradigms
-----------------------------------------------------------------------------------------------------------------
POVZETEK:
Prediction of customer behaviour represents a pivotal task in the retail sector, focusing on identifying customers with a high risk of attrition. We analyse a multi-year dataset from a large Slovenian retailer, encompassing detailed customer demographics, purchasing behaviours, and spending habits, represented as time series. We incorporate Convolutional Neural Networks (CNNs) and apply them on the time series, where the features can be viewed as one-dimensional vectors. Furthermore, we incorporate ResNet-like skip connections, as well as Inception modules with different kernel sizes in our architecture to improve the predictive performance compared to traditional ML techniques.
Seminar bo potekal v živo, s pričetkom ob 16:00 uri v učilnici FAMNIT-VP2.
Vabljeni.
ponedeljek, 24. marec 2025 Fairuz Ishrat BHUIYAN: Comparative Analysis of VR and Traditional Simulators in Anti-Aircraft Training
V ponedeljek, 24. marca 2025, 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: 24. marec 2025 ob 16.00 v FAMNIT-VP2.
------------------------------------------------------
PREDAVATELJICA: Fairuz Ishrat BHUIYAN
------------------------------------------------------
Fairuz Ishrat Bhuiyan is a master’s degree student, born and raised in Helsinki, Finland. She studies Computer Engineering at the University of Turku, and she is specializing in Health Technology. As her minor subject, she studied Marine Technology / Naval Architecture in the Department of Mechanical Engineering at Aalto University, School of Engineering. Her Bachelors thesis topic was Weather and Air Quality in Helsinki and it was part of a national project called Cityzer. In 2021, she completed her one-year military service at the Guard Jaeger Regiment in Santahamina, Helsinki and was promoted to the rank of corporal.
-----------------------------------------------------------------------------------------------------------------------
NASLOV: Comparative Analysis of VR and Traditional Simulators in Anti-Aircraft Training
-----------------------------------------------------------------------------------------------------------------------
POVZETEK:
The study aims to measure attitude processes when using simulators and subjective experiences after use by comparing a VR-based immersive simulator and a classic non-immersive simulator. The target groups are the conscripts in the Parola Armored Brigade of the Finnish Defense Forces, who will receive training in using an anti-aircraft machine gun with the help of virtual reality. Both simulators highlighted commendable traits, such as prompt responsiveness to user actions and awareness of control devices. The VR simulator's propensity to induce more pronounced simulator sickness symptoms occurred similarly in both simulators, suggesting common physiological responses across different simulator types. Interestingly, participants in the VR simulator found it easier to control events and manipulate objects, leading to better engagement than in real-life scenarios. These findings illuminate the advantages and limitations of VR training with augmented cues, offering valuable insights for refinement and future research endeavors.
Seminar bo potekal v živo, s pričetkom ob 16:00 uri v učilnici FAMNIT-VP2.
Vabljeni.
petek, 7. marec 2025 Jan JOVAN: ChatGPT in ostali jezikovni modeli
V ponedeljek, 10. marca 2025, 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. marec 2025 ob 16.00 v FAMNIT-VP2.
-----------------------------------
PREDAVATELJ: Jan JOVAN
-----------------------------------
Jan Jovan je študent 2. letnika magistrskega programa računalništva in informatike ter razvojni inženir v podjetju Smart Com d.o.o.. Specializira se za integracijo velikih jezikovnih modelov (LLM), kar mu omogoča, da prispeva k razvoju naprednih rešitev v podjetju. Njegovo delo vključuje raziskovanje in implementacijo LLM tehnologij ter iskanje novih načinov za izboljšanje njihove uporabnosti v različnih aplikacijah.
------------------------------------------------------------
NASLOV: ChatGPT in ostali jezikovni modeli
------------------------------------------------------------
POVZETEK:
Veliki jezikovni modeli (LLM) so revolucionarizirali obdelavo naravnega jezika z izkoriščanjem obsežnih podatkovnih nizov in arhitektur globokega učenja za generiranjem besedila, odgovarjanjem na vprašanja in izvajanjem kompleksnega sklepanja. Ta seminar raziskuje osnovne mehanizme LLM, vključno z njihovo arhitekturo in metodami usposabljanja, pri čemer izpostavlja naraščajoče število primerov njihove uporabe na različnih področjih. Konkretno se obravnavajo aplikacije LLM v medicini za pomoč pri diagnozi in generiranju medicinske literature, v robotiki za izboljšanje interakcije med človekom in robotom ter v kodiranju za avtomatizacijo razvoja programske opreme. Poudarjeni so tudi izzivi, povezani z ocenjevanjem modelov, vključno z obsegom znanja in razširljivostjo, ter potencialne prihodnje smeri za izboljšanje zmogljivosti LLM.
Seminar bo potekal v živo, s pričetkom ob 16:00 uri v učilnici FAMNIT-VP2.
Vabljeni.
četrtek, 13. februar 2025 Gianmarco CHERCHI: Interactive and Robust Mesh Booleans
V ponedeljek, 17. februarja 2025, 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: 17. februar 2025 ob 16.00 v FAMNIT-VP2.
------------------------------------------------
PREDAVATELJ: Gianmarco CHERCHI
------------------------------------------------
Gianmarco Cherchi is a Computer Science Researcher (Tenure Track Assistant Professor) at the Department of Mathematics and Computer Science of the University of Cagliari (Italy), where he obtained his PhD. His research interests are in Computer Graphics and Geometry Processing, focusing on the generation and optimization of surface and volumetric meshes, and Digital Fabrication. He is also the Professor of the "Data Visualization" (Applied Computer Science and Data Analytics BSc) and "Web Programming" (Computer Science BSc) courses at the University of Cagliari. In 2024, he received the "Young Investigator Award 2024," granted by the Shape Modeling International Organization.
------------------------------------------------------------------
NASLOV: Interactive and Robust Mesh Booleans
------------------------------------------------------------------
POVZETEK:
Boolean operations are among the most used paradigms to create and edit digital shapes. Despite being conceptually simple, the computation of mesh Booleans is notoriously challenging. The main issues come from numerical approximations that make the detection and processing of intersection points inconsistent and unreliable, exposing implementations based on floating-point arithmetic to many kinds of failure. Numerical methods based on rational numbers or exact geometric predicates have the needed robustness guarantees that are achieved at the cost of increased computation times that, as of today, have always restricted the use of robust mesh Booleans to offline applications. In this seminar, I will briefly summarize the results obtained in three recent articles on the topic, which have enabled us to develop an algorithm for Boolean operations with robustness guarantees, capable of operating at interactive frame rates on meshes with up to 200K triangles.
Seminar bo potekal v živo, s pričetkom ob 16:00 uri v učilnici FAMNIT-VP2.
Vabljeni.
četrtek, 16. januar 2025 Ivan DAMNJANOVIĆ: Finding the number of inequivalent arithmetic expressions on n variables
V ponedeljek, 20. januarja 2025, 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: 20. januar 2025 ob 17.00 v FAMNIT-VP2.
-----------------------------------------------
PREDAVATELJ: Ivan DAMNJANOVIĆ
-----------------------------------------------
Ivan Damnjanović is pursuing a PhD degree in Mathematical Sciences at the Faculty of Mathematics, Natural Sciences and Information Technologies at the University of Primorska. He previously obtained a PhD degree in Electrical Engineering and Computing at the Faculty of Electronic Engineering at the University of Niš, where he currently works as a teaching assistant at the department of mathematics.
-----------------------------------------------------------------------------------------------------------------
NASLOV: Finding the number of inequivalent arithmetic expressions on n variables
-----------------------------------------------------------------------------------------------------------------
POVZETEK:
Given n distinct formal variables, in how many ways can we use them to construct different arithmetic expressions? An expression tree is a rooted tree whose internal nodes correspond to some operations to be performed, while its leaves are formal variables. Here, we deal with the expression trees such that the only allowed operations are the four standard arithmetic operations (addition, subtraction, multiplication and division) together with, optionally, additive inversion. We consider two expressions to be equivalent if their expression trees yield the same formal expression. To begin, we provide certain theoretical results concerning the equivalence of arithmetic expressions. Afterwards, we disclose a Θ(n^2) algorithm that computes the number of inequivalent arithmetic expressions on n distinct variables. The algorithm covers both the case when the unary operation of additive inversion is allowed and when it is not.
(This is a joint work with Ivan Stošić and Žarko Ranđelović.)
Seminar bo potekal v živo, s pričetkom ob 17:00 uri v učilnici FAMNIT-VP2.
Pozor, to je eno uro kasneje kot ponavadi !!!
Vabljeni.
petek, 10. januar 2025 Jovan PAVLOVIĆ: A Data-Driven Approach for the Analysis of Ridership Fluctuations in Transit Systems
V ponedeljek, 13. januarja 2025, 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: 13. januar 2025 ob 16.00 v FAMNIT-VP2.
------------------------------------------
PREDAVATELJ: Jovan PAVLOVIĆ
------------------------------------------
Jovan Pavlović is a second-year Data Science master's student at UP FAMNIT, where he also earned his bachelor's degree in Mathematics.
--------------------------------------------------------------------------------------------------------------------------------
NASLOV: A Data-Driven Approach for the Analysis of Ridership Fluctuations in Transit Systems
--------------------------------------------------------------------------------------------------------------------------------
POVZETEK:
This seminar explores a data-driven approach to analyzing ridership fluctuations in public transportation systems, especially during pandemics. It focuses on identifying critical components within urban transit systems, by employing agent-based simulations and graph analytics techniques. Key findings reveal specific transit stops and routes that are highly sensitive to changes in demand, often serving as bottlenecks or high-risk areas for the spread of infectious diseases.
Seminar bo potekal v živo, s pričetkom ob 16:00 uri v učilnici FAMNIT-VP2.
Vabljeni.
