Scholarlink

This paper extends previous research results and proposes a fine-tuning experimental analysis of a Bayesian-Neural-Networks (BNN) framework that focuses the attention on scaling-posterior-distribution problems over heterogeneous curated datasets that are accessed, processed and managed by different users. In particular, we employ state-of-the-art image datasets and competitive metrics to fine-tuning evaluate the capabilities of our proposed framework, specifically in comparison with competitive approaches and techniques. Retrieved results clearly demonstrate the benefits deriving from our research, by reaching the best performance with a PSNR value of 47.36529 dB and a SSIM value of 0.95578 \(\%\) , thus confirming its relevance for modern intelligent big data analytics applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Fine-Tuning Experimental Analysis of a Bayesian-Neural-Networks Framework Over Heterogeneous Curated Datasets

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Alfredo Cuzzocrea

摘要

This paper extends previous research results and proposes a fine-tuning experimental analysis of a Bayesian-Neural-Networks (BNN) framework that focuses the attention on scaling-posterior-distribution problems over heterogeneous curated datasets that are accessed, processed and managed by different users. In particular, we employ state-of-the-art image datasets and competitive metrics to fine-tuning evaluate the capabilities of our proposed framework, specifically in comparison with competitive approaches and techniques. Retrieved results clearly demonstrate the benefits deriving from our research, by reaching the best performance with a PSNR value of 47.36529 dB and a SSIM value of 0.95578 \(\%\) , thus confirming its relevance for modern intelligent big data analytics applications.