Performance analysis of association rules algorithms on automotive industry data with spmf and weka

Data Mining is a branch of science that enables the analysis of existing data on any data set to make meaningful inferences or to predict future data with technical methods. This contributes to the development of computer-aided decision-making mechanisms based on predictions or inferences. With the rapidly developing technology, companies serving in the wholesale and retail sector can now store their data much faster, easier and with lower costs. All transactions performed during the day (sales, current card, invoicing, etc.) in the companies combine at the end of the day to form big data sets. It is possible to derive some useful inferences both for companies and customers from these data sets which are rapidly increasing in size. At this stage, data mining is used to make the inferences mentioned. In this study, Turkey's many regions of car care products to sell at a company- owned data set, Data Mining Market Basket Analysis Association Rule algorithms latest 11 algorithm is applied and the rules of the products made in conjunction sale have been identified. Thanks to these rules, it is possible to redefine sales and marketing strategies for the related company, to revise the storage areas efficiently, and to create sales campaigns suitable for customers and regions. In this thesis, Apriori and FP-Growth algorithms, which are the two most commonly used algorithms, were run separately for different support values in both WEKA and SPMF and the performance values of both programs were compared graphically. After the SPMF was found to be more successful than WEKA, the operations were continued with this software and the working time of the 11 current association rules algorithms on the relevant data set, the total memory used during the run, the number of rules issued for the relevant algorithms were calculated in the SPMF program. the inferences were compared graphically for different support values. As a result of the application performed in SPMF software, dEclat_bitset algorithm showed the most efficient performance for 6 months and 12 months dataset. However, it can be said that Eclat algorithm is the most efficient algorithm for support values of 0.7 and 0.3 in the 22-month dataset; on the other hand, dEclat_bitset is the most efficient algorithm for support values of 0.3 and 0.1 in the 22-month dataset.