20.500.12556/RUL-172931 Uporaba strojnega učenja v rastlinski genetiki Use of machine learning in plant genetics for comprehension of plants in the digital era Namen diplomske naloge je bil preučiti uporabo metod strojnega učenja v genetiki rastlin za poglobitev našega razumevanja rastlinskega sveta. Z boljšim razumevanjem rastlin in kompleksnih lastnosti kot sta povezava med genotipom in fenotipom bi se lahko spopadli s perečimi okoljskimi izzivi, ki se pojavljajo v agronomiji. Ugotovila sem, da metode strojnega učenja omogočajo učinkovito obdelavo velikih količin podatkov, povečujejo natančnost napovedi fenotipov in optimizacijo že obstoječih tehnologij kot je npr. editiranje genoma. Številne raziskave so uspešno uporabile metode strojnega učenja za doseganje svojih rezultatov, s tem pa so dokazale perspektivnost teh tehnologij za prihodnost. Kljub temu pa ostajajo pomembne omejitve, kot so pomanjkanje standardiziranih podatkov, omejena interpretabilnost modelov in težave pri prenosu rezultatov iz laboratorijskih v naravne pogoje. Rezultati kažejo, da je za nadaljni napredek potrebna tesnejša integracija biološkega znanja z računalniškimi pristopi in razvoj jasno interpretabilnih modelov. Za popolnoma uspešno in učinkovito uporabo v agronomiji so potrebne še nadaljne raziskave, predvsem pa vztrajno nadaljno dopolnjevanje podatkovnih zbirk. The aim of this thesis was to investigate the application of machine learning methods in plant genetics in order to enhance our understanding of the plant kingdom. By better understanding plants and complex traits such as the relationship between genotype and phenotype, we could address pressing environmental challenges in agronomy. I found that machine learning methods enable efficient processing of large amounts of data, increase the accuracy of phenotype prediction, and optimize existing technologies such as genome editing. Numerous studies have successfully applied machine learning methods to achieve their results, demonstrating the promising future of these technologies. Nevertheless, significant limitations remain, such as the lack of standardized data, limited interpretability of models, and challenges in transferring results from laboratory to natural conditions. The findings indicate that further progress requires closer integration of biological knowledge with computational approaches and the development of clearly interpretable models. For fully successful and effective application in agronomy, further research is necessary, especially consistent expansion of data repositories. rastlinska genetika strojno učenje genotip fenotip aplikacije plant genetics machine learning genotype phenotype aplications true false false Slovenski jezik Angleški jezik Diplomsko delo/naloga 2025-09-12 07:19:20 2025-09-12 07:19:21 2025-10-28 03:51:21 0000-00-00 00:00:00 2025 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 263670 249000707 51785.pdf 51785.pdf 1 34FC2906E1E3222CBC7B58BCAB520687 6a344168d018416cfab0df238bbe561438afcc8de5beaffd7a049dc689d9ba42 df991a89-8f97-11f0-9328-0050569b8976 https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=217065 Biotehniška fakulteta 0 0 0