Search

Medical artificial intelligence (AI) applies AI technology to the medical field to meet clinical needs, combines with the experience of doctors, and relies on the computing, analysis and decision-making ability of AI to provide accurate intelligent assistance for clinical diagnosis and treatment. At present, medical AI is based on knowledge guided and data driven approaches, but each has its own advantages and disadvantages. Combining knowledge -guided AI with data- driven AI and utilizing their respective advantages is expected to break through the application bottle neck of medical AI and promote the development and innovation of medical AI. This paper summarizes the application of medical AI driven by knowledge guidance and data mining and looks forward to the future development direction, in order to promote the innovation and application of AI technology in the medical field. [ABSTRACT FROM AUTHOR]

Concerning the dissimilarity between original spatial data and progressively transmitted spatial data, this paper proposes a model for measuring the similarity of curve datasets. After curve similarity is measured based on differences of the topological and geometric characteristics, the weighted similarity and the similarity precision for curve datasets are computed. The model uses weighted similarity to assess the overall degree of similarity for curve datasets and uses similarity precision to assess the similarity differences between curves. Finally, the model was used in an experiment and verified for effectiveness. [ABSTRACT FROM AUTHOR]

In recent years, the privacy-preserving data mining has become a hotspot in data mining. However, with the development of technologies (mobile communication, embedded and positioning technologies, etc),the emerging applications( the Internet of things, location-based services, social network based on location, etc) result in accumulation of abundant personal privacy information, which will easily lead to the violence of personal privacy. Therefore, it is significant to study the privacy-preserving data mining methods to meet the demands of new applications. In view of the analyzing the characteristics and catalogs of existing privacy-preserving data mining methods, this paper proposed their challenges from the field of new distributed system, high dimensional data and spatio-temporal data, etc, as well as indicated the future research directions. [ABSTRACT FROM AUTHOR]