Scholarlink

In this chapter, we introduce deep learning networks and then consider several of the representational structures that support building these networks. We begin by describing the research of Yoshua Bengio, Geoffrey Hinton, and Yann LeCun that enabled the development of deep networks. We then introduce four early research projects that demonstrate both the successes and the promise of these networks. In ► Sect. 17.2, we describe several meta-parameters used in building these networks, including the use of softmax that converts network output values into a distribution of real numbers between 0 and 1.

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

Deep Learning: Introduction and Representations

  • 章节

  • 发表日期：

    2025年

查看全文
  • George F. Luger

摘要

In this chapter, we introduce deep learning networks and then consider several of the representational structures that support building these networks. We begin by describing the research of Yoshua Bengio, Geoffrey Hinton, and Yann LeCun that enabled the development of deep networks. We then introduce four early research projects that demonstrate both the successes and the promise of these networks. In ► Sect. 17.2, we describe several meta-parameters used in building these networks, including the use of softmax that converts network output values into a distribution of real numbers between 0 and 1.