Application Of Decision Tree Improvement Algorithm In Cost Control Of Small And Medium Enterprises

At present,in most industries,enterprise products are becoming more and more homogenous in terms of performance and quality.At this time,the level of cost management determines the competitiveness of enterprises in the market.In other words,if the company can do cost control,it will become more competitive.The cost management of an enterprise often involves many factors.The effect of cost control depends on these.All the factors run through the whole life cycle of the production or operation of the enterprise.Then,how to obtain the success of the business through data mining It is crucial to look at the elements and find the intrinsic link between these feature characteristics and corporate cost control.In many business management processes,a large amount of information and data are retained.The deep mining and application of these data can often provide a basis for cost management and an important means to implement enterprise cost control.Data mining technology can better achieve the goal of cost control and complete the vertical extension from data representation to data internal relationship.Based on the collection of the cost accounting of a chemical fertilizer enterprise in Shenzhen,this paper uses data mining technology to process the enterprise cost data,and studies the application of the decision tree algorithm in the cost control of the enterprise.The main research contents are as follows:(1)Theoretical overview: This paper gives a theoretical overview of the theoretical data mining technology of decision tree algorithm,and then discusses the classical algorithm of decision tree to lay the theoretical foundation of the full text research.(2)Cost modeling: This paper conducts modeling analysis of decision tree algorithm in enterprise cost control,focusing on the defects of enterprise cost control decision-making,taking cost control as the basic control object,and implementing the decision tree ID3 algorithm by means of Python language.Cost control modeling ultimately builds a series of rules that predict business costs.(3)Algorithm improvement: The improvement of the algorithm is based on the shortcomings and defects of ID3 traditional decision tree algorithm.Through literature research,it is found that there is a multi-value orientation in the selection of decision attributes of traditional ID3 algorithm.PCA's decision tree optimization algorithm,in the ordinary PCA-based decision tree improvement algorithm,there is a problem that the data is not highly representative after the dimension reduction processing,resulting in the algorithm needs to go through multiple data runs to improve theaccuracy.Based on the ID3 algorithm,the attribute feature values ??are extracted twice before classification,and the amount of data to be classified is calculated,that is,the most important attribute selection is performed on the original data.After the subtree is established,the data is further reduced.The combination is selected and the improved algorithm is verified by using three data sets in the UCI database.The results show that the average accuracy rate in the three data sets is 94.6%,which is improved by the traditional ID3 algorithm and the ordinary PCA decision tree optimization algorithm.1.6% and 0.6%,PCA-based decision tree algorithm can improve the accuracy of results to a certain extent,with Given practical significance,ID3 algorithm after the second follow-up will improve the implementation of modeling,to obtain a more efficient decision tree model.(4)Improved analysis: This paper applies the improved ID3 algorithm to enterprise cost control to complete the actual verification.Through research,it is found that the optimized algorithm has obvious superiority to the traditional ID3 algorithm in terms of modeling time,total number of nodes,number of leaf nodes,number of decision rules,and average accuracy.Construction is more practical.In this paper,the decision tree algorithm is applied in the cost control of enterprises.The decision tree algorithm is used to realize the modeling and forecasting of enterprise cost control.Using the forecasting model,the cost control of the enterprise can be targeted and the efficiency of enterprise cost control can be improved.effect.The enterprise cost control model realized by the decision tree algorithm informs the enterprise's cost control process,realizes the sharing and transmission of enterprise historical data between different stages and departments,and provides important data support for cost decision makers to effectively make up The shortcomings of the enterprise in the cost control process.