Analysis of Bayesian optimization algorithms for big data classification based on Map Reduce framework

• Main Authors: Chitrakant Banchhor, N. Srinivasu
• Format: Article
• Language: English
• Published: SpringerOpen 2021-06-01
• Series: Journal of Big Data
• Subjects: Map Reduce; Correlative naive Bayes classifier; Classification; Big data; Holoentropy
• Online Access: https://doi.org/10.1186/s40537-021-00464-4

Abstract The process of big data handling refers to the efficient management of storage and processing of a very large volume of data. The data in a structured and unstructured format require a specific approach for overall handling. The classifiers analyzed in this paper are correlative naïve Bayes classifier (CNB), Cuckoo Grey wolf CNB (CGCNB), Fuzzy CNB (FCNB), and Holoentropy CNB (HCNB). These classifiers are based on the Bayesian principle and work accordingly. The CNB is developed by extending the standard naïve Bayes classifier with applied correlation among the attributes to become a dependent hypothesis. The cuckoo search and grey wolf optimization algorithms are integrated with the CNB classifier, and significant performance improvement is achieved. The resulting classifier is called a cuckoo grey wolf correlative naïve Bayes classifier (CGCNB). Also, the performance of the FCNB and HCNB classifiers are analyzed with CNB and CGCNB by considering accuracy, sensitivity, specificity, memory, and execution time.