This section presents the basic principles and types of an emerging field of science called artificial intelligence. In addition to the term artificial intelligence coined by John McCarthy almost 70 years ago, today the terms machine learning, artificial neural networks, neural networks, deep learning, computer vision, and natural language processing have been added to this field and are used. It is therefore necessary to present these terms and to clarify them. The utilization of the aforementioned methods requires the development of appropriate architectural structures, which are discussed in this section. Since most of the state-of-the-art artificial intelligence technologies in dermatology are based on neural networks (deep learning type), the basic principles underlying them and their mode of operation are presented. This is followed by an analysis of more specialized deep learning technologies, such as convolutional neural networks and their variants, which are state-of-the-art technology in the processing and analysis of dermatological images. The utilization of the aforementioned methods necessitates the development of appropriate architectural structures, which are analyzed in the present section. Given that most state-of-the-art artificial intelligence technologies in dermatology are based on neural networks (a type of deep learning), a presentation of their fundamental principles and operational mechanisms is provided. Subsequently, a more specialized analysis of deep learning technologies follows, such as convolutional neural networks and their variants, which represent cutting-edge technology in the processing and analysis of dermatological images. Two equally important technologies with applications in the field of dermatology are large language models (LLMs) and vision transformers (ViTs), which are also included in this section.

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

Artificial Intelligence (AI) Main Concepts: A Glimpse inside “A.I. Black Box”

  • Dimitrios Mantzaris,
  • Emmanouil Karampinis

摘要

This section presents the basic principles and types of an emerging field of science called artificial intelligence. In addition to the term artificial intelligence coined by John McCarthy almost 70 years ago, today the terms machine learning, artificial neural networks, neural networks, deep learning, computer vision, and natural language processing have been added to this field and are used. It is therefore necessary to present these terms and to clarify them. The utilization of the aforementioned methods requires the development of appropriate architectural structures, which are discussed in this section. Since most of the state-of-the-art artificial intelligence technologies in dermatology are based on neural networks (deep learning type), the basic principles underlying them and their mode of operation are presented. This is followed by an analysis of more specialized deep learning technologies, such as convolutional neural networks and their variants, which are state-of-the-art technology in the processing and analysis of dermatological images. The utilization of the aforementioned methods necessitates the development of appropriate architectural structures, which are analyzed in the present section. Given that most state-of-the-art artificial intelligence technologies in dermatology are based on neural networks (a type of deep learning), a presentation of their fundamental principles and operational mechanisms is provided. Subsequently, a more specialized analysis of deep learning technologies follows, such as convolutional neural networks and their variants, which represent cutting-edge technology in the processing and analysis of dermatological images. Two equally important technologies with applications in the field of dermatology are large language models (LLMs) and vision transformers (ViTs), which are also included in this section.