Shranjeno v:
| Glavni avtor: | |
|---|---|
| Format: | Recurso digital |
| Jezik: | angleščina |
| Izdano: |
Zenodo
2026
|
| Teme: | |
| Online dostop: | https://doi.org/10.5281/zenodo.19394577 |
| Oznake: |
Označite
Brez oznak, prvi označite!
|
Kazalo:
- <p>Abstract: Traditional image processing systems that used rule-based algorithms and hand-defined features during a<br>long time have been used to extract and reason on visual information. Though they generate computationally efficient<br>and transparent decision making with limited adaptability, these lack adaptability resulting in a performance constraint<br>to the tasks in complex, dynamic environments. The more recent Artificial Intelligence (AI) and Decision Intelligence<br>(DI) concepts proposed data-driven models that integrate machine learning, deep learning, and reasoning in context to<br>improve the accuracy of decisions. In this paper, Decision Intelligence powered by AI is compared to traditional<br>decision-making systems based on image processing regarding methodology, performance, scalability, and general<br>applicability in different fields like healthcare diagnosis, surveillance, and industrial control. The paper points out the<br>fact that the two methods present complementary aspects and present a hybrid model that inherits the interpretative<br>abilities of conventional methods alongside the abilities to predict of AI-fueled DI to create a powerful real-world<br>decision-making model.</p>