Administracija predavanja

Predavač: Izv.prof.dr.sc. Tomislav Staroveški i prof.dr.sc. Danko Brezak
Institucija: Fakultet strojarstva i brodogradnje, Sveučilište u Zagrebu
Naslov predavanja (HR): Innovations in Robotic Sanding and Polishing Systems
Naslov predavanja (EN): Innovations in Robotic Sanding and Polishing Systems
Tip predavanja: Seminar za optimizaciju i primjene
Datum i time: 4.03.2026., 12:00
Predavaonica: Predavaonica 2
Sažetak (HR): Advances in robotic sanding and polishing have led to significant improvements in surface finishing quality, consistency, and overall process efficiency. A key challenge in these operations is ensuring precise control over tool engagement while maintaining desired surface characteristics. To address this, recent developments have focused on process monitoring, real-time force control, and adaptive path planning techniques that enhance performance and reliability of sanding or polishing processes. This presentation will cover state-of-the-art methods in robotic sanding and polishing, focusing on force control, and direct and indirect process monitoring approaches aimed for tool condition and surface quality estimation in real-time. Automated path planning strategies, including pre-programmed and adaptive trajectory generation, will be discussed in the context of optimizing tool interaction and ensuring uniform surface quality. Presentation will also address the development of ARCOPS robotic cell, a specialized system for robotic sanding and polishing that integrates both direct and indirect process monitoring subsystems. Key system capabilities, case studies, and real-world applications will be presented.

Sažetak (EN): Advances in robotic sanding and polishing have led to significant improvements in surface finishing quality, consistency, and overall process efficiency. A key challenge in these operations is ensuring precise control over tool engagement while maintaining desired surface characteristics. To address this, recent developments have focused on process monitoring, real-time force control, and adaptive path planning techniques that enhance performance and reliability of sanding or polishing processes. This presentation will cover state-of-the-art methods in robotic sanding and polishing, focusing on force control, and direct and indirect process monitoring approaches aimed for tool condition and surface quality estimation in real-time. Automated path planning strategies, including pre-programmed and adaptive trajectory generation, will be discussed in the context of optimizing tool interaction and ensuring uniform surface quality. Presentation will also address the development of ARCOPS robotic cell, a specialized system for robotic sanding and polishing that integrates both direct and indirect process monitoring subsystems. Key system capabilities, case studies, and real-world applications will be presented.

Predavač: Tomislav Kovačević
Institucija: Fakultet elektrotehnike i računarstva, Sveučilište u Zagrebu
Naslov predavanja (HR): Optimal Trend Labeling in Financial Time Series
Naslov predavanja (EN): Optimal Trend Labeling in Financial Time Series
Tip predavanja: Seminar za optimizaciju i primjene
Datum i time: 11.03.2026., 12:00
Predavaonica: Predavaonica 2
Sažetak (HR): Predicting asset price trends is often posed as a classification problem, where trends are classified as positive or negative. Since asset price series are noisy and volatile, it is difficult to distinguish true trends from short-term fluctuations. To this end, several trend definitions have been proposed in the literature, but it is yet to be known how these trend definitions affect the performance of classification algorithms designed to learn such labels from historical data. In this paper, we define the robustness of the trend labeling algorithm as a measure of how well a classifier designed to learn such labels can withstand a change in the cumulative return considering the classifier’s generalization error. Moreover, we propose a noise model to simulate the desired accuracy score, which allows us to evaluate the robustness of a trend labeling algorithm without the need to train an actual classifier and consequently choose the optimal algorithm in terms of robustness. Experimental results confirm the adequacy of the proposed noise model and show that classification algorithms perform better when trained with such optimal labels.

Sažetak (EN): Predicting asset price trends is often posed as a classification problem, where trends are classified as positive or negative. Since asset price series are noisy and volatile, it is difficult to distinguish true trends from short-term fluctuations. To this end, several trend definitions have been proposed in the literature, but it is yet to be known how these trend definitions affect the performance of classification algorithms designed to learn such labels from historical data. In this paper, we define the robustness of the trend labeling algorithm as a measure of how well a classifier designed to learn such labels can withstand a change in the cumulative return considering the classifier’s generalization error. Moreover, we propose a noise model to simulate the desired accuracy score, which allows us to evaluate the robustness of a trend labeling algorithm without the need to train an actual classifier and consequently choose the optimal algorithm in terms of robustness. Experimental results confirm the adequacy of the proposed noise model and show that classification algorithms perform better when trained with such optimal labels.

Predavač: Davide Palitta
Institucija: Alma Mater Studiorum, Università di Bologna, Italija
Tip predavanja: Seminar za optimizaciju i primjene
Datum i time: 27.05.2026., 12:00
Predavaonica: Predavaonica 2

Predavač: Prof. dr. sc. Slobodan Filipovski
Institucija: Department of Mathematics, University of Primorska, Koper
Naslov predavanja (HR): Naslov će bit naknadno objavljen
Tip predavanja: Matematički kolokvij
Datum i time: 11.06.2026., 14:00
Predavaonica: Predavaonica 3