Baking Industry Consumption Of Data Mining And Its Application Research

With the continuous development of China’s food industry, thebaking industry rapidly grows and has huge market potential. It hasbecome a new growth point of the food processing industry. Meanwhile,not only the workload of customer consumption data analysis of bakingenterprises, but also the difficulty coefficient increases sharply, thetraditional analytical methods have been difficult to meet thedevelopment needs of the industry. There is an urgent need to improve theanalytical tools and methods. Data mining technology is an effectivemethod which can fully mining and analyze the hidden information in thedata and provide decision support for baking enterprise management.Specifically, the research work of this paper includes the followingaspects:First，put forward a kind of improved K-Means algorithm, theshortcomings of the K-means algorithm is improved in this paper. TheK-means algorithm combines with the genetic algorithm (GA) and makesfull use of the advantages of the two, which solves the problem of thetraditional K-means algorithm about determining the clustering number,selecting the initial center and solving noise data sensitive issues. Iris andKDD CUP99dataset demonstrate that the GK-means algorithm has betterperformance than the traditional K-means algorithm, and as a basis forbuilding the baking industry customer segmentation model, thensubdivide customer groups based on customer consumption informationand analyze the characteristics of each type;Second，improves Apriori algorithm to increase efficiency when searching for frequent item sets. KDD CUP99dataset demonstrate thatthe improved Apriori algorithm has higher efficiency than the traditionalApriori algorithm. Use the baking industry customer consumptioninformation as data mining objects, and identify the association rulesbetween the various products of the baking industry. Then according tothe rules, formulate cross-selling strategy.