Research And Implementation Of The Small And Medium-sized Technology Enterprises In Risk Assessment Methods

With the continuous development of the system of China’s reform and opening up,China’s comprehensive national strength and economic strength improve continuously.Therefore, some of the emerging enterprise forms have appeared, the proportion of small andmedium-sized technology enterprises in the new economic landscape possession has becamemore and more larger, many experts and scholars focus on the study of small andmedium-sized technology enterprises, the risk assessment is one of the important branches.Nowadays, although the moment of small and medium-sized technology enterprisesdevelops rapidly, its own organizational structure destined to lead its management which hasrisks, so risk management is particularly important for enterprise. So far, there are many riskassessment methods for enterprise, some of which have also been applied, whereas the riskassessment method for the small and medium-sized technology enterprises rarelyconsidered.Thus, the risk assessment for small and medium-sized technology enterprises hasmore research field and value.The author reads a lot of literatures on Risk Management, Support Vector Machinetechnology, Data Mining, Genetic Algorithm and so on. Because of the basis of the work,conclusion the risk assessment methods for small and medium-sized technology enterprises,especially for optimizing Multi-Class Classification Support Vector Machines based onBinary Tree Kernel Function Parameters.The main work of this paper can be divided into the following two aspects:The research problem is the risk of the small and medium-sized technology enterprises,so the first part of this article is the analysis of small and medium-sized technologyenterprises. According to the organizational structure of the small and medium-sizedtechnology enterprises, analysising a full range of the internal characteristics of the small andmedium-sized technology enterprises; estimating the prospects of small and medium-sizedtechnology enterprises; dividing the possible risks of small and medium-sized technologyenterprises; analysising these risk categories, and the presence of hazardous, considering thecharacteristics of existing risk assessment methodologies applicable conditions, applied thesemethods to small and medium-sized technology enterprises, through the corresponding effectof these risk assessment methods in the small and medium-sized technology enterprises,analysis the advantages and disadvantages of these assessment methods comprehensivly.The contents of this study is mainly based on Support Vector Machine technology, Support Vector Machine technology is a new technology in the field of data mining in recentyears, its main idea is that the statistical theory of VC dimension theory, first classificatingand coding the risk of the small and medium-sized technology enterprises, then by theGA-based feature selection algorithm to choose many of the risk factors, derived the riskcategory optimal feature subset, using multi-class Support Vector Machines based on binarytree algorithm for training optimal feature subset, by testing category indicators to optimizerisk assessment model. when the risk assessment model sets up, I use the ant colony algorithmoptimized kernel function parameters, thus ensuring the accuracy of risk assessment usingsupport vector machine for small and medium-sized technology enterprises, and more value inuse.