Table 1. The summary of related work done in stream data analysis.

Sr. no	Author (Year)	Title	Topic/Focus/Purpose	Paradigm and method	Pros or cons
1	L. Du, Q. Song, and X. Jia (2014)	Detecting concept drift: An information entropy-based method using an adaptive sliding window	Window Based Drift Detection Algorithm	Sliding Window Technique for concept drift detection, Window size determined dynamically, entropy is used for drift detection	Cons: determination of window size is difficult
2	J. Gama, P. Medas, G. Castillo, and P. Rodrigues (2014)	Learning with drift detection	Statistical (Error rate based) drift detection method	DDM, used error rate for drift detection, change in error rate determines the change in distribution,	Cons: problems with imbalanced data
3	S. Wang, L. L. Minku, D. Ghezzi, D. Caltabiano, P. Tino, and X. Yao (2013)	Concept drift detection for online class imbalance learning	Statistical (minority class recall-based) Drift detection method	DDM-OCI, considers minority class recall for drift detection, works well with imbalanced data,	Pros: good with imbalanced data, cons: minority and majority class should be known already
4	Heng Wang and Zubin Abraham (2015)	Concept drift detection for streaming data	Statistical (Linear four rates) Drift detection method	Linear Four Rates (LFR) considers tpr, tnr, ppv and npv for drift detection,	Pros: works good with imbalanced data, works fine when imbalance data changes in balance data, cons: Problem of false positive, very sensitive to any change in data distribution and noise
5	Shujian Yu et al. (2019)	Concept drift detection and adaptation with hierarchical hypothesis testing	Two-layer hypothesis test-based drift detection method	upgraded Linear Four Rates to Hierarchical Linear Four Rates, using two-layer architecture tries to minimize the false positive	Pros: minimizes the false positive drift detection Cons: multiple tests each time
6	Dariusz Brzezinski and Jerzy Stefanowski (2011)	Accuracy updated ensemble for data streams with concept drift	Accuracy Based ensemble classifier for drift handling	Accuracy Updated Ensemble, select classifier with Hoeffding Option Tree, update classifiers with current distribution,	Cons: Limited diversity between classifiers in ensemble
7	Iman Khamassi et al. (2013)	Ensemble classifiers for drift detection and monitoring in dynamic environments	Ensemble classifier for concept drift detection and monitoring	Combines online bagging and boosting with EDIST drift detection algorithm,	Cons: EDIST is an error distance-based algorithm, doesn’t work well with imbalanced data
8	Zeng Li et al. (2019)	Drift-detection Based Incremental Ensemble for Reacting to Different Kinds of Concept Drift	Drift detection-based ensemble classifier	Combines DDM, a drift detection method with an incremental ensemble. Works fine with sudden drift.	Cons: DDM is an error-based method, doesn’t work well with imbalanced data.