Survey on Fuzzy Associative Classifications Techniques and Their Performance Evaluation with Different Fuzzy Clustering Techniques Over Big Data

摘要

In the current trend of data science research lot of works concentrated over scaling conventional data classification algorithms towards handling massive datasets referred as Big data. The standard of classification algorithms evaluated based on their accuracy, interpretability and intuitiveness. It is well proved in studies that the using fuzzy generalization in classification will improves the above three factors of classification model. Considering the importance of fuzzy associative classification models this article revise their inception to current extension to MapReduce frameworks. The primary challenge in such techniques is in deciding which clustering technique is well suited for extracting fuzzy generalization parameters. Considering that an experimental study conducted on various clustering algorithms extended to MapReduce framework and their performance evaluated based on effectiveness of the generated fuzzy parameters with respect to classification accuracy.