Fusion Of Symbols And Connections

Taking the research of machine ethics decision support system as the main line,with the help of different research methods of symbolism and connectionism,research on the history of robot decision-making,such as how problems are generated,what are the specific algorithm forms,what models are established,and The research status at home and abroad,the existing solution ideas are sorted and analyzed,and based on this,try to solve the robot decision problem.The general structure of this article is divided into three parts.The first part sorts out the existing results of neural network research,the second part sorts out the main ideas of the symbolic algorithm model,and the third part tries to combine the two methods to try an effective model fusion process..The main content is concentrated in chapter 234.A simple integration of the processes experienced in the basic theoretical stage of the neural network is made,and then a detailed analysis of the MP formula,the perceptron principle,the BP model and the basic formulas of the Dropout idea,the form and proof of the inference,and the various forms Comparative analysis of the versions.In chapter three,for G?del's theorem and formalizable systems,Leibniz's idea of symbolic logic is analyzed and explained.Bayesian induction is used to prove mathematical probability inference systems.In chapter 4,this article starts with the original text of the literature research,and analyzes the philosophical,semantic,and formal aspects of the AMA system,and analyzes the reasons and results of the imperfect decision support system of machine ethics.And try to come up with a solution on this basis.