Concept Drift Adaptation Methods under the Deep Learning Framework: A Literature Review
Abstract
:1. Introduction
- (1)
- We review concept drift adaptation methods under the deep learning framework from four aspects—discriminative learning, generative learning, hybrid learning, and relevant others—so as to fill the gap in this area of investigation in previous work.
- (2)
- We reveal the general operation process of concept drift adaptive methods under deep learning frameworks and explain concept drift detection modes and update modes in detail.
- (3)
- We summarize the representative algorithms of each subcategory, common datasets, evaluation metrics, their application areas, and limitations.
- (4)
- We analyze and discuss the current problems of concept drift adaption methods and point out the future direction.
2. Overview of Concept Drift
2.1. The Definition of Concept Drift
2.2. The Causes of Concept Drift
- (1)
- Virtual concept drift. When the probability of x changes, but the probability of y under the condition of x does not change, i.e., Pt0(x) ≠ Pt1(x) and Pt0(y|x) = Pt1(y|x). This case belongs to virtual concept drift, which does not affect its decision boundary and only changes the feature space.
- (2)
- Real concept drift. When the probability of y under the condition of x changes, the probability of x remains the same, i.e., Pt0 (y|x) ≠ Pt1 (y|x) and Pt0 (x) = Pt1 (x). This case has a direct impact on the prediction model and is a real concept drift, which not only changes the feature space but also changes its decision-making boundary.
- (3)
- Hybrid concept drift. In an open environment, both real concept drift and virtual concept drift can exist in the data stream at the same time, i.e., Pt0 (x) ≠ Pt1 (x), Pt0 (y|x) ≠ Pt1 (y|x). This is a mixed concept drift, which is most common.
2.3. The Types of Concept Drift
2.4. The Process of Concept Drift Adaptation Methods under Deep Learning Framework
3. Concept Drift Adaptation Methods under Deep Learning
3.1. Concept Drift Adaptation Methods Based on Discriminant Learning
- MLP-based concept drift adaptation methods
- RNN-based concept drift adaptation methods
- LSTM-based concept drift adaptation methods
- CNN-based concept drift adaptation methods
3.2. Concept Drift Adaptation Methods Based on Generative Learning
- AE-based concept drift adaptation methods
- GAN-based concept drift adaptation methods
- RBM-based concept drift adaptation methods
- SOM-based concept drift adaptation methods
3.3. Concept Drift Adaptation Methods Based on Hybrid Learning
Types of Deep Learning | Algorithms | Concept Drift Adaptation Methods | Limitation | ||
---|---|---|---|---|---|
Detection Modes | Update Modes | Adaptation Drift Types | |||
LSTM + SNN | HSN-LSTM [71] | − | √ | N | High resource overhead |
LSTM + AE + ORA | OAR-DLSTM [72] | − | √ | I R | The dataset is too large, and its performance appears to degrade |
LSTM + AE | B-Detection [73] | − | √ | A I G R | Long running time |
LSTM + CNN | CausalConvLSTM [74] | − | √ | N | Limited log types for algorithm applications |
LSTMCNNcda [75] | + | √ | A G | Time-series data normalization issues, window size selection | |
AE + DNN | SAE-DNN [76] | + | × | A G | Noise interference |
RNN + ARIMA | OARIMA-RNN [39] | − | √ | N | No quantification of conceptual drift or performance during drift |
MLP + Decision tree + SVM | RACE [77] | + | √ | A G R | Requires large amounts of memory, increased integration size, slows convergence |
3.4. Other Concept Drift Adaptation Methods
- DTL-based concept drift adaptation methods
- DRL-based concept drift adaptation methods
Types of Deep Learning | Algorithms | Concept Drift Adaptation Methods | Limitation | ||
---|---|---|---|---|---|
Detection Modes | Update Modes | Adaptation Drift Types | |||
DTL | NN-Patching [82] | − | × | N | Need to adjust hyperparameters for the scene |
AM-CNNs [83] | + | × | N | High overhead | |
ATL [84] | − | × | A I G | Forgetting problem | |
ADTCD [85] | − | √ | A I G R | Little attention is paid to scarce anomaly data | |
DRL | DeepPocket [87] | − | √ | N | Not suitable for long-term investment strategies |
RL4OASD [88] | − | √ | N | Long model training time | |
OEA-RL [89] | + | √ | N | Updating delay | |
DeepBreath [6] | − | √ | N | Lack of consideration for exogenous factors | |
DFL | FedHAR [90] | − | √ | N | Scarcity of labels, with privacy |
3.5. Discussion
4. Performance Evaluation of Concept Drift
4.1. Datasets
4.2. The Evaluation Metrics
5. Future Directions
5.1. Full Coverage of Concept Drift Types
5.2. Data Processing Problem
5.3. Multi-Model Integration Problem
5.4. Visualization Problem of Concept Drift
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
Abbreviations
OCDD | one-class drift detector |
CDT_MSW | concept drift type identification method based on multi-sliding windows |
KSWIN | Kolmogorov–Smirnov test detector |
LD3 | label dependency drift detector |
STUDD | student–teacher approach for unsupervised drift detection |
CDCMS | concept drift handling based on clustering in the model space |
DMDDM | diversity measure drift detection method |
I-LSTM | improved long short-term memory |
DDM | drift detection method |
MLP | multilayer perceptron |
CNN | convolutional neural network |
RNN | recurrent neural network |
DNN | deep neural network |
SEOA | selective ensemble-based online adaptive deep neural network |
BODL | bilevel online deep learning framework |
NADINE | neural network with dynamically evolved capacity |
CIDD-ADODNN | Adadelta optimizer-based deep neural networks with concept drift detection |
ADWIN | adaptive sliding-window drift detection technology |
OARNN | online adaptive recurrent neural network |
TPE | tree-structured Parzen estimator |
ONU-SHO | opposition-based novel updating spotted hyena optimization |
ONU-SHO-RNN | ONU-SHO-based RNN |
AIBL-MVD | adaptive behavioral-based incremental batch learning malware variant detection model |
SPC | statistical process control |
MUSE-RNN | multilayer self-evolving recurrent neural network |
LSTM | long short-term memory |
CI | continuous integration |
GA | genetic algorithm |
MOMBD-CDD | multi-objective metaheuristic optimization-based big data analytics with concept drift detection |
STEPD | Statistical Test of Equal Proportions method |
GSO | glowworm swarm optimization |
Bi-LSTM | bidirectional long short-term memory |
CUSUM | cumulative sum |
EWMA | exponentially weighted moving average |
AD-LSTM | adaptive LSTM framework |
SDWIN | sliding-window algorithm |
TP-ALS | two-phase adaptive learning strategy |
ECNN | evolutive convolutional neural network |
OS-PGSM | online CNN-based model selection using performance gradient-based saliency maps |
DIH | deep incremental hashing |
ROC | region of competence |
DRT | data reduction technique |
ARCUS | adaptive framework for online deep anomaly detection under a complex evolving data stream |
USCDD-AE | unsupervised statistical concept drift detection |
DEVDAN | deep evolving denoising autoencoder |
MemStream | memory-based streaming anomaly detection |
FIFO | first in, first out |
ADTCD | adaptive anomaly detection approach toward concept drift |
GAN | generative adversarial network |
AE | autoencoder |
RBM | restricted Boltzmann machine |
SOM | self-organizing mapping |
DCIGAN | distributed class-incremental learning method based on generative adversarial networks |
GF | generative fusion |
GRBM | Gaussian restricted Boltzmann machine |
OUIM-SOM | online unsupervised incremental method based on self-organizing maps |
SOINN+ | self-organizing incremental neural network for unsupervised learning from noisy data streams |
AHS | novel adaptive and hybrid spiking module |
OAR-DLSTM | online autoregression with deep long short-term memory |
HDL-OKW | hybrid deep learning classifier and optimized key windowing approach |
BPTT | backpropagation through time |
SAE-DNN | stacked autoencoder-deep neural network |
RVFL | random vector function linking |
SVM | support vector machines |
OARIMA-RNN | adaptive online ensemble learning with recurrent neural network and ARIMA |
RACE | recurrent adaptive classifier ensemble |
DTL | deep transfer learning |
NN-Patching | neural network patching |
AM-CNNs | adaptive mechanism for learning CNNs |
ATL | autonomous transfer learning |
AGMM | autonomous Gaussian mixture model |
DRL | deep reinforcement learning |
OEA-RL | online ensemble aggregation using reinforcement learning |
PH | Page–Hinkley |
KDD | knowledge discovery and data mining |
MCC | Matthews’ correlation coefficient |
MAE | mean absolute error |
MSE | mean squared error |
RMSE | root mean squared error |
HN | the number of hidden nodes per time step |
HL | the number of hidden layers per time step |
PC | parameter count |
ET | execution time |
MDD | maximum drawdown |
Sr | Sharpe ratio |
CVaR | conditional value at risk |
FedHAR | federated learning human activity recognition frame |
References
- Shrestha, Y.R.; Krishna, V.; von Krogh, G. Augmenting organizational decision-making with deep learning algorithms: Principles, promises, and challenges. J. Bus. Res. 2021, 123, 588–603. [Google Scholar] [CrossRef]
- Schlimmer, J.C.; Granger, R.H. Incremental learning from noisy data. Mach. Learn. 1986, 1, 317–354. [Google Scholar] [CrossRef]
- Lu, J.; Liu, A.; Song, Y.; Zhang, G. Data-driven decision support under concept drift in streamed big data. Complex Intell. Syst. 2020, 6, 157–163. [Google Scholar] [CrossRef]
- Gama, J.; Žliobaitė, I.; Bifet, A.; Pechenizkiy, M.; Bouchachia, A. A survey on concept drift adaptation. ACM Comput. Surv. 2014, 46, 1–37. [Google Scholar] [CrossRef]
- Xu, R.; Cheng, Y.; Liu, Z.; Xie, Y.; Yang, Y. Improved Long Short-Term Memory based anomaly detection with concept drift adaptive method for supporting IoT services. Futur. Gener. Comput. Syst. 2020, 112, 228–242. [Google Scholar] [CrossRef]
- Soleymani, F.; Paquet, E. Financial portfolio optimization with online deep reinforcement learning and restricted stacked autoencoder—DeepBreath. Expert Syst. Appl. 2020, 156, 113456. [Google Scholar] [CrossRef]
- Wang, X.; Chen, W.; Xia, J.; Chen, Z.; Xu, D.; Wu, X.; Xu, M.; Schreck, T. ConceptExplorer: Visual Analysis of Concept Drifts in Multi-source Time-series Data. In Proceedings of the IEEE Conference on Visual Analytics Science and Technology (VAST), Salt Lake City, UT, USA, 25–30 October 2020; pp. 1–11. [Google Scholar] [CrossRef]
- Hoens, T.R.; Polikar, R.; Chawla, N.V. Learning from streaming data with concept drift and imbalance: An overview. Prog. Artif. Intell. 2012, 1, 89–101. [Google Scholar] [CrossRef]
- Wen, Y.; Qiang, B.; Fan, Z. A survey of the classification of data streams with concept drift. CAAI Trans. Intell. Syst. 2013, 8, 95–104. [Google Scholar]
- Ditzler, G.; Roveri, M.; Alippi, C.; Polikar, R. Learning in Nonstationary Environments: A Survey. IEEE Comput. Intell. Mag. 2015, 10, 12–25. [Google Scholar] [CrossRef]
- Webb, G.I.; Hyde, R.; Cao, H.; Nguyen, H.L.; Petitjean, F. Characterizing concept drift. Data Min. Knowl. Discov. 2016, 30, 964–994. [Google Scholar] [CrossRef]
- Krawczyk, B.; Minku, L.L.; Gama, J.; Stefanowski, J.; Woźniak, M. Ensemble learning for data stream analysis: A survey. Inf. Fusion 2017, 37, 132–156. [Google Scholar] [CrossRef]
- Khamassi, I.; Sayed-Mouchaweh, M.; Hammami, M.; Ghédira, K. Discussion and review on evolving data streams and concept drift adapting. Evol. Syst. 2018, 9, 1–23. [Google Scholar] [CrossRef]
- Lu, J.; Liu, A.; Dong, F.; Gu, F.; Gama, J.; Zhang, G. Learning under Concept Drift: A Review. IEEE Trans. Knowl. Data Eng. 2018, 31, 2346–2363. [Google Scholar] [CrossRef]
- Ab Ghani, N.L.; Aziz, I.A.; Mehat, M. Concept Drift Detection on Unlabeled Data Streams: A Systematic Literature Review. In Proceedings of the IEEE Conference on Big Data and Analytics (ICBDA), Xiamen, China, 8–11 May 2020; pp. 61–65. [Google Scholar] [CrossRef]
- Lima, M.; Neto, M.; Filho, T.S.; Fagundes, R.A.D.A. Learning Under Concept Drift for Regression—A Systematic Literature Review. IEEE Access 2022, 10, 45410–45429. [Google Scholar] [CrossRef]
- Bayram, F.; Ahmed, B.S.; Kassler, A. From concept drift to model degradation: An overview on performance-aware drift detectors. Knowl. Based Syst. 2022, 245, 108632. [Google Scholar] [CrossRef]
- Desale, K.S.; Shinde, S.V. Addressing Concept Drifts Using Deep Learning for Heart Disease Prediction: A Review. In Proceedings of Second Doctoral Symposium on Computational Intelligence; Gupta, D., Khanna, A., Kansal, V., Fortino, G., Hassanien, A.E., Eds.; Springer: Singapore, 2022; Volume 1374, pp. 157–167. [Google Scholar] [CrossRef]
- Iwashita, A.S.; Papa, J.P. An Overview on Concept Drift Learning. IEEE Access 2018, 7, 1532–1547. [Google Scholar] [CrossRef]
- R.O. Duda, P.E. Hart, and D.G. Stork, Pattern Classification, New York: John Wiley & Sons, 2001, pp. xx + 654, ISBN: 0-471-05669-3. J. Classif. 2007, 24, 305–307. [CrossRef]
- Agrahari, S.; Singh, A.K. Concept Drift Detection in Data Stream Mining: A literature review. J. King Saud Univ.-Comput. Inf. Sci. 2021, 34, 9523–9540. [Google Scholar] [CrossRef]
- Korycki, Ł.; Krawczyk, B. Adversarial concept drift detection under poisoning attacks for robust data stream mining. Mach. Learn. 2022, 1–36. [Google Scholar] [CrossRef]
- Gözüaçık, Ö.; Can, F. Concept learning using one-class classifiers for implicit drift detection in evolving data streams. Artif. Intell. Rev. 2021, 54, 3725–3747. [Google Scholar] [CrossRef]
- Guo, H.; Li, H.; Ren, Q.; Wang, W. Concept drift type identification based on multi-sliding windows. Inf. Sci. 2022, 585, 1–23. [Google Scholar] [CrossRef]
- Raab, C.; Heusinger, M.; Schleif, F.-M. Reactive Soft Prototype Computing for Concept Drift Streams. Neurocomputing 2020, 416, 340–351. [Google Scholar] [CrossRef]
- Gözüaçık, Ö.; Büyükçakır, A.; Bonab, H.; Can, F. Unsupervised concept drift detection with a discriminative classifier. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, Beijing China, 3–7 November 2019; pp. 2365–2368. [Google Scholar]
- Cerqueira, V.; Gomes, H.M.; Bifet, A.; Torgo, L. STUDD: A student–teacher method for unsupervised concept drift detection. Mach. Learn. 2022, 1–28. [Google Scholar] [CrossRef]
- Chiu, C.W.; Minku, L.L. A Diversity Framework for Dealing with Multiple Types of Concept Drift Based on Clustering in the Model Space. IEEE Trans. Neural Netw. Learn. Syst. 2020, 33, 1299–1309. [Google Scholar] [CrossRef]
- Mahdi, O.A.; Pardede, E.; Ali, N.; Cao, J. Diversity measure as a new drift detection method in data streaming. Knowl.-Based Syst. 2020, 191, 105227. [Google Scholar] [CrossRef]
- Yuan, L.; Li, H.; Xia, B.; Gao, C.; Liu, M.; Yuan, W.; You, X. Recent Advances in Concept Drift Adaptation Methods for Deep Learning. In Proceedings of the 31st International Joint Conference on Artificial Intelligence, Vienna, Austria, 23–29 July 2022; International Joint Conferences on Artificial Intelligence Organization: Vienna, Austria, 2022; pp. 5654–5661. [Google Scholar]
- Sarker, I.H. Deep Learning: A Comprehensive Overview on Techniques, Taxonomy, Applications and Research Directions. SN Comput. Sci. 2021, 2, 420. [Google Scholar] [CrossRef]
- Leväsluoto, J.; Kohl, J.; Sigfrids, A.; Pihlajamäki, J.; Martikainen, J. Digitalization as an Engine for Change? Building a Vision Pathway towards a Sustainable Health Care System by Using the MLP and Health Economic Decision Modelling. Sustainability 2021, 13, 13007. [Google Scholar] [CrossRef]
- Guo, H.; Zhang, S.; Wang, W. Selective ensemble-based online adaptive deep neural networks for streaming data with concept drift. Neural Netw. 2021, 142, 437–456. [Google Scholar] [CrossRef]
- Han, Y.-N.; Liu, J.-W.; Xiao, B.-B.; Wang, X.-T.; Luo, X.-L. Bilevel Online Deep Learning in Non-stationary Environment. In Proceedings of the 30th International Conference on Artificial Neural Networks, Bratislava, Slovakia, 14–17 September 2021. [Google Scholar] [CrossRef]
- Pratama, M.; Za’In, C.; Ashfahani, A.; Ong, Y.S.; Ding, W. Automatic Construction of Multi-layer Perceptron Network from Streaming Examples. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, Beijing China, 3–7 November 2019. [Google Scholar] [CrossRef]
- Priya, S.; Uthra, R.A. Deep learning framework for handling concept drift and class imbalanced complex decision-making on streaming data. Complex Intell. Syst. 2021, 1–17. [Google Scholar] [CrossRef]
- Dezfouli, A.; Griffiths, K.; Ramos, F.; Dayan, P.; Balleine, B.W. Models that learn how humans learn: The case of decision-making and its disorders. PLoS Comput. Biol. 2019, 15, e1006903. [Google Scholar] [CrossRef]
- Fekri, M.N.; Patel, H.; Grolinger, K.; Sharma, V. Deep learning for load forecasting with smart meter data: Online Adaptive Recurrent Neural Network. Appl. Energy 2021, 282, 116177. [Google Scholar] [CrossRef]
- Jagait, R.K.; Fekri, M.N.; Grolinger, K.; Mir, S. Load Forecasting Under Concept Drift: Online Ensemble Learning with Recurrent Neural Network and ARIMA. IEEE Access 2021, 9, 98992–99008. [Google Scholar] [CrossRef]
- Singh, M.N.; Khaiyum, S. Enhanced Data Stream Classification by Optimized Weight Updated Meta-learning: Continuous learning-based on Concept-Drift. Int. J. Web Inf. Syst. 2021, 17, 645–668. [Google Scholar] [CrossRef]
- Darem, A.A.; Ghaleb, F.A.; Al-Hashmi, A.A.; Abawajy, J.H.; Alanazi, S.M.; Al-Rezami, A.Y. An Adaptive Behavioral-Based Incremental Batch Learning Malware Variants Detection Model Using Concept Drift Detection and Sequential Deep Learning. IEEE Access 2021, 9, 97180–97196. [Google Scholar] [CrossRef]
- Das, M.; Pratama, M.; Savitri, S.; Zhang, J. MUSE-RNN: A Multilayer Self-Evolving Recurrent Neural Network for Data Stream Classification. In Proceedings of the IEEE International Conference on Data Mining (ICDM), Beijing, China, 8–11 November 2019; pp. 110–119. [Google Scholar] [CrossRef]
- Badri, A.K.; Heikal, J.; Terah, Y.A.; Nurjaman, D.R. Decision-Making Techniques using LSTM on Antam Mining Shares before and during the COVID-19 Pandemic in Indonesia. Aptisi Trans. Manag. 2021, 6, 167–180. [Google Scholar] [CrossRef]
- Saidani, I.; Ouni, A.; Mkaouer, M.W. Improving the prediction of continuous integration build failures using deep learning. Autom. Softw. Eng. 2022, 29, 21. [Google Scholar] [CrossRef]
- Mansour, R.F.; Al-Otaibi, S.; Al-Rasheed, A.; Aljuaid, H.; Pustokhina, I.V.; Pustokhin, D.A. An Optimal Big Data Analytics with Concept Drift Detection on High-Dimensional Streaming Data. Comput. Mater. Contin. 2021, 68, 2843–2858. [Google Scholar] [CrossRef]
- Alencar, B.M.; Canário, J.P.; Neto, R.L.; Prazeres, C.; Bifet, A.; Rios, R.A. Fog-DeepStream: A new approach combining LSTM and Concept Drift for data stream analytics on Fog computing. Internet Things 2023, 22, 100731. [Google Scholar] [CrossRef]
- Luo, X.; Zhang, D. An adaptive deep learning framework for day-ahead forecasting of photovoltaic power generation. Sustain. Energy Technol. Assess. 2022, 52, 16. [Google Scholar] [CrossRef]
- Diez-Olivan, A.; Ortego, P.; Del Ser, J.; Landa-Torres, I.; Galar, D.; Camacho, D.; Sierra, B. Adaptive Dendritic Cell-Deep Learning Approach for Industrial Prognosis Under Changing Conditions. IEEE Trans. Ind. Inform. 2021, 17, 7760–7770. [Google Scholar] [CrossRef]
- Cheng, M.-M.; Jiang, P.-T.; Han, L.-H.; Wang, L.; Torr, P. Deeply Explain CNN Via Hierarchical Decomposition. Int. J. Comput. Vis. 2023, 131, 1091–1105. [Google Scholar] [CrossRef]
- Gu, J.; Wang, Z.; Kuen, J.; Ma, L.; Shahroudy, A.; Shuai, B.; Liu, T.; Wang, X.; Wang, G.; Cai, J.; et al. Recent advances in convolutional neural networks. Pattern Recognit. 2018, 77, 354–377. [Google Scholar] [CrossRef]
- Li, X.; Zhang, Z.; Zhao, Z.; Wu, L.; Huo, J.; Zhang, J.; Wang, Y. ECNN: One Online Deep Learning Model for Streaming Ocean Data Prediction. In Proceedings of the ACM International Conference on Intelligent Computing and Its Emerging Applications, Jinan, China, 28–29 December 2021; p. 6. [Google Scholar]
- Saadallah, A.; Jakobs, M.; Morik, K. Explainable Online Deep Neural Network Selection Using Adaptive Saliency Maps for Time Series Forecasting. In Machine Learning and Knowledge Discovery in Databases. Research Track; Oliver, N., Pérez-Cruz, F., Kramer, S., Read, J., Lozano, J.A., Eds.; Springer International Publishing: Cham, Switzerland, 2021; Volume 12975, pp. 404–420. [Google Scholar] [CrossRef]
- Tian, X.; Ng, W.W.Y.; Xu, H. Deep Incremental Hashing for Semantic Image Retrieval with Concept Drift. IEEE Trans. Big Data 2023, 1–13. [Google Scholar] [CrossRef]
- Lobo, J.L.; Laña, I.; Del Ser, J.; Bilbao, M.N.; Kasabov, N. Evolving Spiking Neural Networks for online learning over drifting data streams. Neural Netw. 2018, 108, 1–19. [Google Scholar] [CrossRef]
- Da’u, A.; Salim, N. Recommendation system based on deep learning methods: A systematic review and new directions. Artif. Intell. Rev. 2020, 53, 2709–2748. [Google Scholar] [CrossRef]
- Deng, L. A tutorial survey of architectures, algorithms, and applications for deep learning. APSIPA Trans. Signal Inf. Process. 2014, 3, e2. [Google Scholar] [CrossRef]
- Bank, D.; Koenigstein, N.; Giryes, R. Autoencoders. arXiv 2021, arXiv:2003.05991. [Google Scholar]
- Yoon, S.; Lee, Y.; Lee, J.-G.; Lee, B.S. Adaptive Model Pooling for Online Deep Anomaly Detection from a Complex Evolving Data Stream. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Washington, DC, USA, 14–18 August 2022; pp. 2347–2357. [Google Scholar] [CrossRef]
- Friedrich, B.; Sawabe, T.; Hein, A. Unsupervised statistical concept drift detection for behaviour abnormality detection. Appl. Intell. 2022, 53, 2527–2537. [Google Scholar] [CrossRef]
- Ashfahani, A.; Pratama, M.; Lughofer, E.; Ong, Y.-S. DEVDAN: Deep evolving denoising autoencoder. Neurocomputing 2020, 390, 297–314. [Google Scholar] [CrossRef]
- Bhatia, S.; Jain, A.; Srivastava, S.; Kawaguchi, K.; Hooi, B. MemStream: Memory-Based Streaming Anomaly Detection. In Proceedings of the ACM Web Conference 2022, Lyon, France, 25–29 April 2022; pp. 610–621. [Google Scholar] [CrossRef]
- Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative adversarial networks. Commun. ACM 2020, 63, 139–144. [Google Scholar] [CrossRef]
- Guan, H.; Wang, Y.; Ma, X.; Li, Y. DCIGAN: A Distributed Class-Incremental Learning Method Based on Generative Adversarial Networks. In Proceedings of the IEEE Intl Conf on Parallel & Distributed Processing with Applications, Big Data & Cloud Computing, Sustainable Computing & Communications, Social Computing & Networking (ISPA/BDCloud/SocialCom/SustainCom), Rio de Janeiro, Brazil, 20–24 May 2019; pp. 768–775. [Google Scholar] [CrossRef]
- Guzy, F.; Wozniak, M.; Krawczyk, B. Evaluating and Explaining Generative Adversarial Networks for Continual Learning under Concept Drift. In Proceedings of the International Conference on Data Mining Workshops (ICDMW), Auckland, New Zealand, 7–10 December 2021; pp. 295–303. [Google Scholar] [CrossRef]
- Zhang, N.; Ding, S.; Zhang, J.; Xue, Y. An overview on Restricted Boltzmann Machines. Neurocomputing 2018, 275, 1186–1199. [Google Scholar] [CrossRef]
- Korycki, L.; Krawczyk, B. Concept Drift Detection from Multi-Class Imbalanced Data Streams. In Proceedings of the IEEE 37th International Conference on Data Engineering (ICDE), Chania, Greece, 19–22 April 2021; pp. 1068–1079. [Google Scholar]
- Wang, W.; Zhang, M. Data Stream Adaptive Partitioning of Sliding Window Based on Gaussian Restricted Boltzmann Machine. In Artificial Intelligence in China; Liang, Q., Wang, W., Mu, J., Liu, X., Na, Z., Chen, B., Eds.; Springer: Singapore, 2020; Volume 572, pp. 220–228. [Google Scholar] [CrossRef]
- Miljkovic, D. Brief review of self-organizing maps. In Proceedings of the 40th International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO), Opatija, Croatia, 22–26 May 2017; pp. 1061–1066. [Google Scholar]
- Cerri, R.; Junior, J.D.C.; Faria, E.R.; Gama, J. A new self-organizing map based algorithm for multi-label stream classification. In Proceedings of the 36th Annual ACM Symposium on Applied Computing, Virtual Event Republic of Korea, 22–26 March 2021. [Google Scholar] [CrossRef]
- Wiwatcharakoses, C.; Berrar, D. SOINN+, a Self-Organizing Incremental Neural Network for Unsupervised Learning from Noisy Data Streams. Expert Syst. Appl. 2020, 143, 113069. [Google Scholar] [CrossRef]
- Zheng, W.; Zhao, P.; Chen, G.; Zhou, H.; Tian, Y. A Hybrid Spiking Neurons Embedded LSTM Network for Multivariate Time Series Learning under Concept-drift Environment. IEEE Trans. Knowl. Data Eng. 2022, 1–14. [Google Scholar] [CrossRef]
- Sun, L.; Ji, Y.; Zhu, M.; Gu, F.; Dai, F.; Li, K. A new predictive method supporting streaming data with hybrid recurring concept drifts in process industry. Comput. Ind. Eng. 2021, 161, 107625. [Google Scholar] [CrossRef]
- Wang, L.; Chen, S.; Chen, F.; He, Q.; Liu, J. B-Detection: Runtime Reliability Anomaly Detection for MEC Services with Boosting LSTM Autoencoder. IEEE Trans. Mob. Comput. 2023, 1–14. [Google Scholar] [CrossRef]
- Yen, S.; Moh, M.; Moh, T.-S. CausalConvLSTM: Semi-Supervised Log Anomaly Detection Through Sequence Modeling. In Proceedings of the 18th IEEE International Conference on Machine Learning And Applications (ICMLA), Boca Raton, FL, USA, 16–19 December 2019; pp. 1334–1341. [Google Scholar]
- Perera, P.P.B. Anomaly Detection on Single Variate Time Series with Concept drifts. Ph.D. Thesis, University of Colombo School of Computing, Colombo, Sri Lanka, 2021. [Google Scholar]
- Zhang, X.; Zou, Y.; Li, S. Enhancing incremental deep learning for FCCU end-point quality prediction. Inf. Sci. 2020, 530, 95–107. [Google Scholar] [CrossRef]
- Museba, T.; Nelwamondo, F.; Ouahada, K.; Akinola, A. Recurrent Adaptive Classifier Ensemble for Handling Recurring Concept Drifts. Appl. Comput. Intell. Soft Comput. 2021, 2021, 5533777. [Google Scholar] [CrossRef]
- Massaoudi, M.; Chihi, I.; Abu-Rub, H.; Refaat, S.S.; Oueslati, F.S. Convergence of Photovoltaic Power Forecasting and Deep Learning: State-of-Art Review. IEEE Access 2021, 9, 136593–136615. [Google Scholar] [CrossRef]
- Mnih, V.; Kavukcuoglu, K.; Silver, D.; Rusu, A.A.; Veness, J.; Bellemare, M.G.; Graves, A.; Riedmiller, M.; Fidjeland, A.K.; Ostrovski, G.; et al. Human-level control through deep reinforcement learning. Nature 2015, 518, 529–533. [Google Scholar] [CrossRef]
- Zhang, D.; China Electric Power Research Institute; Han, X.; Deng, C. Taiyuan University of Technology. Review on the research and practice of deep learning and reinforcement learning in smart grids. CSEE J. Power Energy Syst. 2018, 4, 362–370. [Google Scholar] [CrossRef]
- Tan, C.; Sun, F.; Kong, T.; Zhang, W.; Yang, C.; Liu, C. A Survey on Deep Transfer Learning. In Proceedings of the 27th International Conference on Artificial Neural Networks, Rhodes, Greece, 4–7 October 2018. [Google Scholar] [CrossRef]
- Kauschke, S.; Lehmann, D.H.; Furnkranz, J. Patching Deep Neural Networks for Nonstationary Environments. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 14–19 July 2019. [Google Scholar]
- Disabato, S.; Roveri, M. Learning Convolutional Neural Networks in presence of Concept Drift. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 14–19 July 2019; p. 8. [Google Scholar] [CrossRef]
- Pratama, M.; de Carvalho, M.; Xie, R.; Lughofer, E.; Lu, J. ATL: Autonomous Knowledge Transfer from Many Streaming Processes. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, Beijing, China, 3–7 November 2019. [Google Scholar]
- Xu, L.; Ding, X.; Peng, H.; Zhao, D.; Li, X. ADTCD: An Adaptive Anomaly Detection Approach Towards Concept-Drift in IoT. IEEE Internet Things J. 2023, 1–13. [Google Scholar] [CrossRef]
- Yang, T.; Tang, H.; Bai, C.; Liu, J.; Hao, J.; Meng, Z.; Liu, P.; Wang, Z. Exploration in Deep Reinforcement Learning: A Comprehensive Survey. arXiv 2022, arXiv:2109.06668. [Google Scholar]
- Soleymani, F.; Paquet, E. Deep graph convolutional reinforcement learning for financial portfolio management—DeepPocket. Expert Syst. Appl. 2021, 182, 115127. [Google Scholar] [CrossRef]
- Zhang, Q.; Wang, Z.; Long, C.; Huang, C.; Yiu, S.M.; Liu, Y.; Cong, G.; Shi, J. Online Anomalous Subtrajectory Detection on Road Networks with Deep Reinforcement Learning. arXiv 2022, arXiv:2211.08415. [Google Scholar] [CrossRef]
- Saadallah, A.; Morik, K. Online Ensemble Aggregation using Deep Reinforcement Learning for Time Series Forecasting. In Proceedings of the IEEE 8th International Conference on Data Science and Advanced Analytics (DSAA), Porto, Portugal, 6–9 October 2021; pp. 1–8. [Google Scholar] [CrossRef]
- Yu, H.; Chen, Z.; Zhang, X.; Chen, X.; Zhuang, F.; Xiong, H.; Cheng, X. FedHAR: Semi-Supervised Online Learning for Personalized Federated Human Activity Recognition. IEEE Trans. Mob. Comput. 2021, 22, 3318–3332. [Google Scholar] [CrossRef]
- Liu, Z.; Loo, C.K.; Pasupa, K.; Seera, M. Meta-cognitive recurrent kernel online sequential extreme learning machine with kernel adaptive filter for concept drift handling. Eng. Appl. Artif. Intell. 2020, 88, 103327. [Google Scholar] [CrossRef]
- da Silva, C.A.S.; Krohling, R.A. Semi-Supervised Online Elastic Extreme Learning Machine with Forgetting Parameter to deal with concept drift in data streams. In Proceedings of the International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 14–19 July 2019; pp. 1–8. [Google Scholar]
- Tsukada, M.; Kondo, M.; Matsutani, H. A Neural Network-Based On-device Learning Anomaly Detector for Edge Devices. IEEE Trans. Comput. 2020, 69, 1027–1044. [Google Scholar] [CrossRef]
- Asuncion, A.; Newman, D. UCI Machine Learning Repository; University of California: Irvine, CA, USA, 2007. [Google Scholar]
- Harries, M.; Wales, N.S. Splice-2 Comparative Evaluation: Electricity Pricing; Research Gate: Berlin, Germany, 1999. [Google Scholar]
- Elwell, R.; Polikar, R. Incremental Learning of Concept Drift in Nonstationary Environments. IEEE Trans. Neural Netw. 2011, 22, 1517–1531. [Google Scholar] [CrossRef]
- Katakis, I.; Tsoumakas, G.; Vlahavas, I. Tracking recurring contexts using ensemble classifiers: An application to email filtering. Knowl. Inf. Syst. 2010, 22, 371–391. [Google Scholar] [CrossRef]
- Lopez-Paz, D.; Ranzato, M.A. Gradient Episodic Memory for Continual Learning. arXiv 2022, arXiv:1706.08840. [Google Scholar]
- Street, W.N.; Kim, Y. A streaming ensemble algorithm (SEA) for large-scale classification. In Proceedings of the seventh ACM SIGKDD international conference on Knowledge discovery and data mining, San Francisco, CA, USA, 26–29 August 2001; ACM Press: New York, NY, USA, 2001; pp. 377–382. [Google Scholar]
- Gama, J.; Medas, P.; Castillo, G.; Rodrigues, P. Learning with Drift Detection. In Advances in Artificial Intelligence—SBIA 2004; Bazzan, A.L.C., Labidi, S., Eds.; Springer: Berlin/Heidelberg, Germany, 2004; Volume 3171, pp. 286–295. [Google Scholar] [CrossRef]
- Fan, W.; Yu, P.; Han, J. Mining Concept-Drifting Data Streams using Ensemble Classifiers. In Proceedings of the 9th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Washington, DC, USA, 24–27 August 2003; p. 10. [Google Scholar]
- Kakisim, A.G.; Nar, M.; Sogukpinar, I. Metamorphic malware identification using engine-specific patterns based on co-opcode graphs. Comput. Stand. Interfaces 2020, 71, 103443. [Google Scholar] [CrossRef]
- Anguita, D.; Ghio, A.; Oneto, L.; Parra-Llanas, X.; Reyes-Ortiz, J. Proceedings of the International Conference on European Symposium on Artificial Neural Networks (ESANN); ESAN: Bruges, Belgium, 2013; Volume 3, p. 3. [Google Scholar]
- Coates, A.; Ng, A.; Lee, H. An analysis of single-layer networks in unsupervised feature learning. In Proceedings of the 14th International Conference on Artificial Intelligence and Statistics, Fort Lauderdale, FL, USA, 11–13 April 2011; pp. 215–223. [Google Scholar]
- Elson, J.; Douceur, J.R.; Howell, J.; Saul, J. Asirra: A CAPTCHA that exploits interest-aligned manual image categorization. CCS 2007, 7, 366–374. [Google Scholar]
- Krizhevsky, A.; Hinton, G. Learning Multiple Layers of Features from Tiny Images; University of Toronto: Toronto, ON, Canada, 2009. [Google Scholar]
- Xu, W.; Huang, L.; Fox, A.; Patterson, D.; Jordan, M.I. Detecting large-scale system problems by mining console logs. In Proceedings of the ACM SIGOPS 22nd Symposium on Operating Systems Principles, Big Sky, MT, USA, 11–14 October 2009; ACM Press: New York, NY, USA, 2009. [Google Scholar]
- Sharafaldin, I.; Lashkari, A.H.; Ghorbani, A.A. Toward generating a new intrusion detection dataset and intrusion traffic characterization. In Proceedings of the International Conference on Information Systems Security and Privacy, Funchal, Portugal, 22–24 January 2018; pp. 108–116. [Google Scholar]
- Pang, G.; Shen, C.; Cao, L.; Hengel, A.V.D. Deep learning for anomaly detection: A review. ACM Comput. Surv. 2021, 54, 1–38. [Google Scholar] [CrossRef]
- Ganaie, M.A.; Hu, M.; Malik, A.; Tanveer, M.; Suganthan, P. Ensemble deep learning: A review. Eng. Appl. Artif. Intell. 2022, 115, 105151. [Google Scholar] [CrossRef]
- Yang, W.; Li, Z.; Liu, M.; Lu, Y.; Cao, K.; Maciejewski, R.; Liu, S. Diagnosing concept drift with visual analytics. In Proceedings of the IEEE Conference on Visual Analytics Science and Technology (VAST), Salt Lake City, UT, USA, 25–30 October 2020; pp. 12–23. [Google Scholar]
- Stiglic, G.; Kokol, P. Interpretability of Sudden Concept Drift in Medical Informatics Domain. In Proceedings of the IEEE 11th International Conference on Data Mining Workshops, Vancouver, BC, Canada, 11–14 December 2011; pp. 609–613. [Google Scholar] [CrossRef]
Types of Deep Learning | Algorithms | Concept Drift Adaptation Methods | Limitation | ||
---|---|---|---|---|---|
Detection Modes | Update Modes | Adaptation Drift Types | |||
MLP | SEOA [33] | − | √ | A I G | Not suitable for high-dimensional non-linear problems |
BODL [34] | + | √ | A | New classes cannot be identified online | |
NADINE [35] | + | × | A G R | Slow training time | |
CIDD-ADODNN [36] | + | √ | A G | Feature selection to be optimized | |
RNN | OARNN [38] | − | √ | N | Requires large amounts of data to update the model |
ONU-SHO-RNN [40] | + | √ | I G | Model update delay | |
AIBL-MVD [41] | + | √ | I G | Must have a marked malware sample before updating model | |
MUSE-RNN [42] | + | × | A G R | Cannot handle image data streams | |
LSTM | DL-CIBuild [44] | − | × | N | High labor cost and need to build annotated datasets |
I-LSTM [5] | − | √ | G | Balance of old and new data | |
MOMBD-CDD [45] | + | √ | N | High resource cost | |
Fog-DeepStream [46] | + | √ | High memory consumption | ||
AD-LSTM [47] | + | √ | N | Model update latency exists | |
DCA-DNN [48] | + | √ | N | ||
CNN | ECNN [51] | − | √ | N | High computational cost |
OS-PGSM [52] | + | √ | N | Hyperparameter settings need to be optimized | |
DIH [53] | − | √ | I G | No consideration of semantic relationships between labels | |
SNN | OeSNN [54] | + || − | × | G R | No consideration of a priori information such as speed and severity of drift |
Types of Deep Learning | Algorithms | Concept Drift Adaptation Methods | Limitation | ||
---|---|---|---|---|---|
Detection Modes | Update Modes | Adaptation Drift Types | |||
AE | ARCUS [58] | − | √ | A I G R | Cannot store data for the current batch where concept drift may occur |
DEVDAN [60] | + | × | A I G | Ignores mutation forgetting when adding new layers | |
MemStream [61] | − | √ | N | High resource overhead | |
USCDD-AE [59] | + | √ | A I G R | Difficult data collection and possible false positives | |
GAN | DCIGAN [63] | − | √ | N | Hyperparameter setting |
RBM | RRBM–DD [22] | + | √ | A G R | Limitations in identifying adversarial conceptual drift in dynamic class imbalanced data streams |
RBM–IM [66] | + | √ | A I G | Not suitable for small data streams, prone to overfitting | |
GRBM [67] | + | √ | N | Does not adaptively partition heterogeneous data | |
SOM | OUIM-SOM [69] | − | √ | A I | Limited adaptive effect on conceptual drift |
SOINN | SOINN+ [70] | − | × | A I | Euclidean distance used in the node similarity measure is not suitable for high-dimensional data |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Xiang, Q.; Zi, L.; Cong, X.; Wang, Y. Concept Drift Adaptation Methods under the Deep Learning Framework: A Literature Review. Appl. Sci. 2023, 13, 6515. https://doi.org/10.3390/app13116515
Xiang Q, Zi L, Cong X, Wang Y. Concept Drift Adaptation Methods under the Deep Learning Framework: A Literature Review. Applied Sciences. 2023; 13(11):6515. https://doi.org/10.3390/app13116515
Chicago/Turabian StyleXiang, Qiuyan, Lingling Zi, Xin Cong, and Yan Wang. 2023. "Concept Drift Adaptation Methods under the Deep Learning Framework: A Literature Review" Applied Sciences 13, no. 11: 6515. https://doi.org/10.3390/app13116515