ARTIFICIAL INTELLIGENCE Seminar

 

PROGRAM

---

Predavanja možete pratiti i online putem MITEAM stranice Seminara iz veštačke inteligencije:
https://miteam.mi.sanu.ac.rs/asset/DPP9i2jhvYzp8dmRe

---

Plan rada Seminara iz veštačke inteligencije za SEPTEMBER 2026.

---

Sreda, 30.09.2026. u 19:00, Online
Irina Perfilieva, University of Ostrava, IRAFM, Ostrava, Czech Republic; Cracow University of Technology, Faculty CS&M, Krakow, Poland
MACHINE LEARNING AND FUZZY SETS: TOGETHER OR SEPARATE?
The primary focus of machine learning is the analysis of experimental data and all possible generalizations based on it. In this context, generalization is the ability to process and reason about new, previously unencountered objects or properties of data after gaining experience with a training dataset.
While traditional processing varies by domain—like signals or images—effective data-driven generalization should transcend the physical nature of the data. Instead, it must focus on core tasks like classification and prediction by grouping objects based on their features.
In this talk, we present a theoretical solution to the regression problem and demonstrate its practical application to real-world data samples. We introduce a novel approach that bridges fuzzy sets and neural networks through three key contributions:

• Kernel definition: Resolving whether a uniform fuzzy partition defines a positive-definite (p-d) kernel.
• Feature characterization: Evaluating the unique features extracted directly from the fuzzy partition kernel.
• Network initialization: Demonstrating step-by-step how to initialize a neural network using this strategy.

---

Ovaj onlajn seminar nastao je kao nastavak sastanka "Serbian AI Meeting" i zamišljen je da na njemu istraživači iz Srbije i iz dijaspore, kao i istraživači sa univerzteta, naučnih instituta i iz prakse predstavljaju naučne teme i rezultate iz oblasti veštačke inteligencije.
Link za svako pojedinačno predavanje biće dostavljen dan pre održavanja predavanja.

---

---