Foundations of real-world intelligence

摘要: In 1992 Japan's Ministry of International Trade and Industry (MITI) began a research program in Real World Computing, as successor to the Fifth Generation Computing previous decade, complementing fifth-generation approach. Its objective is lay foundation pursue technical realisation humanlike flexible intelligent information processing. This book collects results ten years original by five laboratories Japan Europe, whose focus has been theoretical algorithmic foundations intelligence manifested real world an our dealing with it. Real-world systems handle complex, uncertain, dynamic, multimodal time. Both explicit implicit are important. Hence we need develop novel integrated framework representing knowledge making inferences based it. It impossible pre-program all needed for coping variety complexity environments, therefore learning adaptation keys intelligence. Learning kind meta-programming strategy. Instead writing programs specific tasks, must write that modify themselves on system's interaction its environment. The includes chapters inference graphical models, approximate reasoning, evolutionary computation beyond, methodology distributed active learning, computing large random patterns. The treatment mathematically rigorous, discussion issues general interest educated reader at large. provides excellent reading graduate courses Computer Science, Cognitive Artificial Intelligence, Applied Statistics. Yoshinori Uesaka professor sciences Science University Tokyo. Pentti Kanerva senior researcher Swedish Institute Science. Hideki Asoh Electrotechnical Laboratory Tsukuba City,