Uncertainty-Based Pattern Mining for Maximizing Profit of Manufacturing Plants With List Structure

Abstract

Products in manufacturing plants are not always manufactured without defects. The probability that commodities are produced without defects is uncertain. Uncertainty-based pattern mining can discover information about a set of goods by considering the possibilities. Besides, products have different importance due to the diverse characteristics of goods. Therefore, we propose a list-based pattern mining method over uncertain data considering an important condition in this article. The proposed method extracts commodities with large values that take into account the importance of merchandise and probability that can be as nondefective products. A list structure is efficient to be created and store a database as a minimal expression. The proposed approach is able to find results more accurately and faster than the existing techniques. We compare the performance of our proposed method with those of the state-of-the-art approaches through real datasets and synthetic datasets. Through these performance tests, we prove that the technique presented in this article has a more excellent performance than the latest algorithms in terms of execution time, memory usage, and scalability.