in the past 2017, deep learning technology vigorous development, AlphaZero started its rapid development in a variety of competitive chess from "zero", DeepStack and Libratus to beat the human master poker in Dezhou, GAN derived varieties, speech synthesis to products from the laboratory, Vicarious put forward a new probabilistic generative model and break the human CAPTCHA verification code. These inspiring progress has brought intelligent technology from laboratory to industry and application level. Artificial intelligence and deep learning concepts have also entered the public view and become popular words.
as the first systematic focus on artificial intelligence technology media, deep learning leading another artificial intelligence tide and enthusiasm of public concern to witness the heart of machines in the past few years, reports of work, at the same time we also found that because of "AI Effect", has a lot of ignores other artificial intelligence technology deep learning the historical process and the direction of development of artificial intelligence technology is based on the concept, classification and technique are the lack of a system of understanding, lead to the overall development trend of the artificial intelligence technology and the availability of "lack of macro understanding, and in an attempt to make use of industrial innovation artificial intelligence technology take a lot of detours.
. Therefore, the heart of the machine introduces 1) the system comprehensively surveys the 23 branch technology of the artificial intelligence (AI).
2) clarifying the historical development path of each branch technology under the artificial intelligence (AI), reading the existing bottleneck and future development trend.
3) of artificial intelligence (AI) application under various branches of technology in the industry, evaluate the "research" and "Engineering", "application", "social influence" of the four stages of the location, to provide a reference for the plan to use artificial intelligence technology decision makers.
4) to provide technical reference for the trend of AI , provide systematic technical learning materials.
What are you going to get from the report?
problem solving (search)
knowledge, reasoning, and planning
communication, perception and behavior
we classify the 23 technologies listed above according to their basic mechanisms, methods and application similarity in AI field. These classifications are not mutually exclusive. You can see them as a human ability, each of which has different functions to work together to achieve a more advanced and complex goal.
From: 除了深度学习，你需要知道AI技术的23个方向 | 机器之心首份技术报告
Translate by: Baidu Translate .