Bootstrap aggregative mean shift clustering for big data anti-pattern detection analytics in 5G/6G communication networks

Abstract

Bootstrap Aggregative Mean-Shift SQL Query Clustering (BAMSQLQC) technique is to identify the anti-patterns in the big data logs with a minimal false-positive rate. BAMSQLQC technique collects numbers of patterns (i.e. queries) from input SQL query big data logs and creates bootstrap big data samples by using the patterns in the given dataset. The BAMSQLQC technique constructs several weak clusters for each pattern in a bootstrap sample. The clustering output of all weak mean-shift clustering is combined into a strong cluster by using the voting scheme to efficiently group similar patterns together with a lesser false-positive rate. The BAMSQLQC technique conducts the experimental results using metrics such as anti-patterns detection accuracy, false-positive rate, time and space complexity. The results show that the BAMSQLQC technique can increase the accuracy and reduce the time complexity of anti-patterns discovery for effective big data analytics in 5G networks compared to existing techniques.