EUVIMEDEUVIMED entdecken
Crossref

Explainable Artificial Intelligence (XAI): Improving Transparency In Deep Learning Models

IOSR Journal of Computer Engineering · 2026-01 · 28 (1) · 57-64

DOI: 10.9790/0661-2801025764 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IOSR Journal of Computer Engineering, 2026-01.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The accelerated growth of Deep Learning (DL) has reshaped the future of Artificial Intelligence (AI), making a breakthrough in the performance of applications in both medical and natural language processing. But such performance can often be accompanied by interpretability; contemporary neural networks can be treated as a black box and they make decisions based on millions of parameters, which are incoherent and unintelligible to human stakeholders. Such lack of transparency raises serious legal, ethical, and practical problems especially in such high- stakes sectors as health care, finance, and criminal justice. Explainable Artificial Intelligence (XAI) has become one of the most important disciplines to resolve this tension, and it seeks to generate mechanisms to render the actions of complex models interpretable to humans without causing much adverse predictive power. The paper gives an in-depth 7,000-word discussion of the XAI landscape. We review the theoretical background of interpretability systematically and classify the most advanced methods such as LIME, SHAP and Grad-CAM, and critically analyze the frameworks employed to measure the quality of explanations. In addition, we will examine the legal requirements imposed by such regulations as GDPR and comment on the future of neuro-symbolic solutions. The study will combine existing literature and technical innovations to propose that XAI is not a feature but a compulsory condition to deploy autonomous systems in the society responsibly.

Zitieren

APA

(2026). Explainable Artificial Intelligence (XAI): Improving Transparency In Deep Learning Models. IOSR Journal of Computer Engineering, 28(1), 57-64. https://doi.org/10.9790/0661-2801025764

Vancouver

Explainable Artificial Intelligence (XAI): Improving Transparency In Deep Learning Models. IOSR Journal of Computer Engineering. 2026;28(1):57-64. doi:10.9790/0661-2801025764

BibTeX
@article{euvimed2026,
  title = {Explainable Artificial Intelligence (XAI): Improving Transparency In Deep Learning Models},
  journal = {IOSR Journal of Computer Engineering},
  year = {2026},
  volume = {28},
  number = {1},
  pages = {57--64},
  doi = {10.9790/0661-2801025764}
}
Kostenlos starten →