“The four 150-minute exams, which took 15 minutes and 8 seconds per game, scored as high as 456 points and exceeded the passing score of 96 points.†The artificial intelligence “intellectual assistant†robot jointly developed by Tsinghua University and HKUST News was on the 14th. This is officially released here. From August 26th to 27th this year, "Intelligent Medical Assistant" as an artificial intelligence robot participated in the 2017 National Practitioner's Qualification Examination in the disconnected and unsigned examination room designated by the National Medical Examination Center. After reviewing and reviewing by experts, the “Intelligent Medical Assistant†robot achieved a score of 456 points and was among the 530,000 candidates in the country. The “Intelligent Medical Assistant†robot passed the comprehensive written test in the medical field, and it is the first time in the world. Wu He, director of the joint laboratory of Tsinghua Unicom, told reporters that massive medical knowledge is difficult to express in computers, and traditional knowledge mapping technology cannot meet the needs of describing medical knowledge. In order to solve the learning problem, Wu Hehe and his team proposed a set of “semantic tensor†methods. According to the requirements of medical examinations, the entire field of medical materials, clinical guidelines, and classic cases of medical undergraduates are used to make the machine obtain “quantification of the medical fieldâ€. "The concept of expression and relationship means that the machine has a huge medical knowledge base. In the use of medical knowledge to analyze and solve problems, the research team used a multi-scale fusion inference algorithm including “key point semantic reasoningâ€, “contextual semantic reasoningâ€, and “evidence chain semantic reasoning†to allow machines to have vocabulary and sentences. Multi-level reasoning between paragraphs. "When answering questions, the medical examination robot first analyzes the questions and options first through autonomous thinking. At the same time, it looks for possible evidence of correct answers for each option, and measures the relevance and availability of the assessment questions, options, and evidence. Confidence, and ultimately make comprehensive decisions, come to the answer." Wu and said. According to the introduction of Tao Xiaodong, general manager of China Medical News Medical Technology Co., Ltd., “Intelligent Medical Assistant†robot will be promoted nationwide in March 2018. The robot will act as a small secretary and can be used in clinical decision support, standardized training, medical education, etc. Aspects. Anhui Hefei is building the second comprehensive national science center in China. China's only National Engineering Laboratories for Brain-like Intelligent Technology and Applications and the National Intelligent Voice High-Tech Industrialization Base, which is uniquely positioned for artificial intelligence, are located in Hefei. Source: China News