Journal of King Saud University - Computer and Information Sciences, Journal Year: 2024, Volume and Issue: 36(5), P. 102079 - 102079

Published: May 31, 2024

Deep learning (DL) is one of the most promising technological developments emerging in fourth industrial revolution era for businesses to improve processes, increase efficiency, and reduce errors. Accordingly, hierarchical software selection critical decision-making problems integrating neural network applications into business models. However, selecting appropriate reinforcement deep enterprises' models takes much work decision-makers. There are several reasons this: first, practitioners' limited knowledge experience DL makes it difficult decision-makers adapt this technology their model significantly increases complex uncertainties. Secondly, according authors' knowledge, no study literature addresses structured solutions with help MCDM approaches. Consequently, making inferences concerning criteria that should be considered an evaluation process impossible by considering studies relevant literature. Considering these gaps, presents a novel approach developed authors. It involves combination two new approaches, MAXC (MAXimum Criterion) TODIFFA (the total differential alternative), which were solve current problems. When important advantages considered, associates objective subjective approaches eliminates some limitations methodologies. Besides, has easily followable algorithm without need advanced mathematical practitioners provides highly stable reliable results solving Another novelty determined long-term negotiation part comprehensive fieldwork specialists. conclusions obtained using briefly reviewed, C2 "Data Availability Quality" criterion influential software. The C7 "Time Constraints" follows factor. Remarkably, prior research overlooked correlation between performance Learning platforms quality accessibility data. findings underscore necessity platform developers devise enable operate effectively, notwithstanding availability clean, high-quality, adequate Finally, robustness check carried out test validity proposed confirms accuracy implementing suggested model.

Language: Английский