ERASMUS+ KA171: New Insights from the project
In a recent joint study published in Aquaculture (ScienceDirect), we tackled a crucial yet often overlooked question: how do we measure stress in aquatic animals—particularly when they’re exposed to challenging conditions?
We performed a comprehensive review of existing stress-indicating methods in fish and shellfish, distinguishing between two categories:
- Contact indicators, such as measuring cortisol levels in blood or tissues—accurate but invasive and sometimes harmful.
- Non-contact indicators, like monitoring behavior, changes in water chemistry, or using imaging and telemetry—a gentler approach, better suited for animal welfare.
We propose a new classification system to organize these metrics by how invasive they are and how reliably they signal stress.
Key Findings
- Traditional contact methods (e.g., blood cortisol) remain reliable but stressful and often impractical for large-scale use.
- Emerging non-contact techniques—behavioral monitoring, remote imaging, and water-based indicators—show promise for welfare-friendly stress detection.
- The proposed taxonomy helps aquaculturists balance accuracy with animal wellbeing, guiding them toward the most appropriate stress-measurement strategy for specific situations.
The study is available here.
Read MoreObrana teme doktorske disertacije Tomislava Ivanovskog
Dana 20.4. u 12 sati u učionici O-357 Fakulteta za informatiku i digitalne tehnologije doktorand Tomislav Ivanovski obranio je temu doktorskog rada pod nazivom “Predviđanje zrelosti breskve temeljeno na prirodom inspiriranim metaheuristikama” pred povjerenstvom u sastavu prof. dr. sc. Ivo Ipšić, izv. prof. dr. sc. Marija Brkić Bakarić i prof. Boris Vrdoljak te zornim mentorstvom voditeljice projekta prof. dr. sc. Maje Matetić i komentorstvom suradnika doc. dr. sc. Marka Gulića.
Read MoreObjava još jednog rada na međunarodnoj konferenciji DAAAM 2022
Znanstveni rad “Peach firmness prediction using optimized regression trees models” u koautorstvu doktoranda Tomislava Ivanovskog, suradnika prof. dr. sc. Xiaoshuan Zhanga, prof. dr. sc. Tomislava Jemrića, doc. dr. sc. Marka Gulića i voditeljice projekta prof. dr. sc. Maje Matetić prezentiran je i objavljen u zborniku međunarodne konferencije DAAAM 2022.
Read MoreObjava rada na međunarodnoj konferenciji DAAAM 2022
Znanstveni rad “Machine learning for data analysis in football: A survey of methods and problems” u koautorstvu doktoranda Saše Tokića, suradnika doc. dr. sc. Ante Panjkote i voditeljice projekta prof. dr. sc. Maje Matetić prezentiran je i objavljen u zborniku međunarodne konferencije DAAAM 2022.
Read MoreObjava rada u međunarodnom časopisu Sensors
Znanstveni rad “Assessment of Various Machine Learning Models for Peach Maturity Prediction Using Non-Destructive Sensor Data” u koautorstvu člana projekta i doktoranda Dejana Ljubobratovića, suradnika Marka Vukovića, članice projekta izv. prof. dr. sc. Marije Brkić Bakarić, suradnika prof. dr. sc. Tomislava Jemrića i voditeljice projekta prof. dr. sc. Maje Matetić objavljen je u međunarodnom časopisu Sensors (1424-8220) 22 (2022), 15; 5791, 19 (https://www.mdpi.com/1424-8220/22/15/5791/htm).
Read MoreObjava rada u međunarodnom časopisu Electronics
Znanstveni rad “A Smart Tourism Case Study: Classification of Accommodation Using Machine Learning Models Based on Accommodation Characteristics and Online Guest Reviews” u koautorstvu studentice Nole Čumlievski i članice projekta izv.prof. dr. sc. Marije Brkić Bakarić te voditeljice projekta prof. dr. sc. Maje Matetić objavljen je u međunarodnom časopisu Electronics (Q2).
Read More