Mining Transactional Databases for Frequent and High-Utility Fuzzy Sequential Patterns With Time Intervals

Abstract

The field of data mining is progressing rapidly presenting researchers with many opportunities for research. Sequential pattern mining is a popular and long established technique in data mining which extracts data in the form of sequential patterns satisfying a threshold value such as utility, support, profit or a combination of these. The application of fuzzy theory in sequential pattern mining has also been favored leading to more natural liinguistic representation. Researchers have proposed various hybrid algorithms with fuzzification of any one parameter such as time or quantity. This paper proposes a hybrid fuzzy algorithm for mining of frequent and high utility sequential patterns with fuzzification of both purchase time and purchase quantity parameters, thereby giving more useful fuzzy sequential patterns. Experimental results also prove that the proposed algorithm is better than the existing algorithms. INDEX TERMS Fuzzy theory, high utility mining, maximum measure, sequential pattern mining, timeintervals VOLUME 4, 2016

Demand analysis and technology selection of high-end equipment manufacturing innovation growth based on data mining technology

“…For example, the United States has developed a series of advanced products with "high technology" as the core, such as intelligent robots and laser cutting machines, while Germany has introduced various high-tech materials to replace expensive equipment to produce special-purpose special vehicle lathes or automobile engines. Japan has also developed its own new cars specifically for high-end industries, and has formed an innovative driving mode with HEM as the core and developing countries as the core, However, China mainly focuses on the traditional industry field, and the research on HEM in China is still in the initial stage 5,6 . Domestic scholars have analyzed and discussed from different angles.…”