Research On Soft Computing From The Perspective Of Philosophy

"Soft computing",as opposed to "hard computing"(traditional computing),has become the technical core of modern artificial intelligence and big data research.The main characteristics of "hard computing" are static,strict,accurate and symbolic,but it is no longer suitable for dealing with complex problems in real life,such as automatic driving,face recognition,etc.The main characteristics of "soft computing" are flexibility,randomness,ambiguity and uncertainty.It is a brand new method to deal with the new scientific paradigm with the connotation of uncertainty,non-linearity and time irreversibility and with complex problems as the object,such as dealing with human perception,biological evolution and fuzzy reasoning,etc.In recent decades,soft computing has been first widely applied in the field of science and engineering,attracting the attention of a group of computer scientists,engineers and mathematics theorists at home and abroad.Although the study of soft computing has now entered the expansion phase of development,but the study mostly focus on soft computing algorithm theory research and practical application research.It has not attracted much attention of the humanities and social scientists and philosophers.Even if some scholars are aware of the philosophical investigation of soft computing,these studies are relatively single and scattered,without a systematic philosophical review.Neither the scale nor the depth of their research has reached the level matching the development of soft computing in application research.This paper aims to discuss the "meaning construction","autonomous cognition" and "research methods" in soft computing from the philosophical perspective,based on which the paper analyzes the philosophical implication behind them from the aspects of ontology,epistemology and methodology,and reveals the revolution inspired by soft computing.This paper follows the research idea of "point-line-surface-body".Firstly,the whole paper is divided into five core chapters,which correspond to five "research nodes" respectively: "soft computing","meaning construction","autonomous cognition","research methods" and "philosophical implication".According to the analysis of these nodes,they are intertwined in "soft computing" and become "research center".This paper makes "soft computing development history" as "research vertical line" and makes fuzzy sets,neural algorithm,genetic algorithm and so on as "research horizontal line".A center,two lines and five nodes form a surface,which is the "research surface" with the theme of "soft computing".The in-depth study of this surface on the philosophical revolution constitutes a "research body",forming the research idea of "point-line-plane-body",which is to cross lines with points,to band surfaces with lines,and to construct bodies with surfaces.The main research contents of this paper are as follows:Chapter 1: Overview the definition,development history and ideological core of soft computing.First of all,the ideological process of Professor Zadeh from "creating the term of soft computing" to "clarifying the definition of soft computing" is summarized,and the conceptual connotation of "soft computing" is clearly defined and analyzed.Secondly,the development history of "soft computing" research is analyzed from the perspective of historicism.By tracing the germination,rise,maturity and expansion of soft computing research,the historical development prospect of soft computing thinking and theory is understood.Finally,the theoretical basis of the core algorithms of soft computing are briefly introduced,the "soft" and "hard" characteristics of computing are compared and analyzed,and the "softening strategy" of soft computing is analyzed in detail.Based on this,the basic understanding of the core of soft computing is formed.Chapter 2: Analyze systematical of the basic approach of meaning construction in soft computing: simulated representation.How does soft computing theories relate to the real world? The "meaning world" constructed by soft computing theory has become a bridge between the data world and the real world.Then,how is this "meaning world" constructed? This question becomes the core topic of this chapter.Firstly,the viewpoint that "the fundamental approach of meaning construction in soft computing is simulated representation" is put forward clearly,the connotation of "simulated representation" in soft computing is explained,and the basic structure of "simulated representation" is analyzed.Secondly,the simulated representation approach of soft computing is investigated in detail from three aspects of "syntax","semantic" and "pragmatic".Finally,the "contextual basis" of meaning construction in soft computing is investigated.The process of simulated representation is also the process of contextualization of the algorithm model,and the result of contextualization constructs the computational context,and the simulated representation at the syntax,semantic and pragmatic levels is logically unified in the computational context.Chapter 3: The cognitive limitations of hard computing and the cognitive breakthrough of soft computing: autonomy,as well as the development prospect of autonomous cognition of soft computing cognition under the dimension of "intersubjectivity" are deeply explored.How does soft computing give machines some autonomy to understand the world? How will the future development of soft computing autonomous balance the relationship between the "machine cognitive subject" and the "human subject" for the increasingly powerful autonomous cognitive ability?To answer these questions,this chapter first briefly analyzes the cognitive limitations of hard computing.Secondly,the internal mechanism and basic path of autonomous cognition by soft computing are investigated in detail.Finally,the future development direction of autonomous cognition of soft computing is to realize "man-machine symbiosis" in the dimension of "intersubjectivity".Chapter 4: Analyze the metaphor method,model method and fusion method in soft computing.Firstly,the relationship between concept and metaphor in soft computing,the characteristics and methodological significance of metaphor in soft computing are investigated.Secondly,the characteristics of model methods in soft computing and the complex relationship between model methods and holism,between model methods and reductionism in soft computing are discussed in depth.Finally,the basic principles and forms of fusion method in soft computing are analyzed in detail,and the example of fusion method is introduced,and the system thinking in fusion method is analyzed.By exploring the deep methodological significance of these research methods,it can provide important reference value for the development of other scientific theories.Chapter 5: The ontological foundation,epistemological interpretation and methodological analysis of soft computing are comprehensively discussed.Firstly,the historical evolution from "everything is number" to "everything is algorithm" in philosophy is analyzed,and the ontological standpoint of "computationism" is refuted by soft computing,and the ontological basis of soft computing is clarified: "all things are still things".Secondly,the epistemological foundation of soft computing and hard computing is investigated in detail,and it is pointed out that the fusion of algorithms is an inevitable trend in the future,which objectively requires and promotes the fusion of the two epistemologies.Finally,the characteristics and principles of generative holism of soft computing are emphatically analyzed,revealing its transcendence over reductionism and system holism.Conclusion: As a new theory of computer science and technology,soft computing not only promotes the development of artificial intelligence to the depth at the technical level,but also causes a revolutionary change at the thinking level: the revolution of logic view;the revolution of mathematics view;the revolution of scientific research methods;the revolution of the view of nature;the revolution in machine cognition.Soft computing is a computing method that is developing.Currently,the scientific research on soft computing is in full swing,and the technological and thinking changes brought by it will continue to expand and deepen.There is still a long way to go for the philosophical research on soft computing.