Learning from imbalanced data: open challenges and future directions
Open accessThe DOI matches its Crossref record; 70 of the 70 references it cites were found in a registry or an index.
Abstract
Despite more than two decades of continuous development learning from imbalanced data is still a focus of intense research. Starting as a problem of skewed distributions of binary tasks, this topic evolved way beyond this conception. With the expansion of machine learning and data mining, combined with the arrival of big data era, we have gained a deeper insight into the nature of imbalanced learning, while at the same time facing new emerging challenges. Data-level and algorithm-level methods are constantly being improved and hybrid approaches gain increasing popularity. Recent trends focus on analyzing not only the disproportion between classes, but also other difficulties embedded in the nature of data. New real-life problems motivate researchers to focus on computationally efficient, adaptive and real-time methods. This paper aims at discussing open issues and challenges that need to be addressed to further develop the field of imbalanced learning. Seven vital areas of research in this topic are identified, covering the full spectrum of learning from imbalanced data: classification, regression, clustering, data streams, big data analytics and applications, e.g., in social media and computer vision. This paper provides a discussion and suggestions concerning lines of future research for each of them.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Limited data sample size causes biased prediction due to impartiality and incompleteness in the dataset [64].”
From Artificial intelligence bias in medical system designs: a systematic review (2023) · cites this paper as [64]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“It has been well-documented that machine learning training is severely impaired by imbalanced class sizes (Krawczyk, 2016; Johnson and Khoshgoftaar, 2020).”
From Deep Learning-Based Extreme Heatwave Forecast (2022) · cites this paper as Krawczyk, 2016
Experimental — accuracy not yet measuredIn the citing paper
“On the other hand, there is a shortage of data for large river systems (in HYDRoSWOT), resulting in less training data for the machine learning model (Krawczyk, 2016).”
From Enhancing River Channel Dimension Estimation: A Machine Learning Approach Leveraging the National Water Model, Hydrographic Networks, and Landscape Characteristics (2024) · cites this paper as Krawczyk, 2016
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
SMOTE synthetic minority over-sampling technique
Chawla, N. V., Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002) · Journal of Artificial Intelligence Research · doi:10.1613/jair.953 · Cited by 32,865 · cited 1×
“Therefore, more advanced methods were proposed that try to maintain structures of groups and/or generate new data according to underlying distributions [9].”
MethodHe, H., & Garcia, E. A. (2009) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2008.239 · Cited by 10,359
ADASYN: Adaptive synthetic sampling approach for imbalanced learning
He, H., Bai, Y., Garcia, E. A., & Li, S. (2008) · 2008 IEEE International Joint Conference on Neural Networks (IEEE World Congress on Computational Intelligence) · doi:10.1109/ijcnn.2008.4633969 · Cited by 4,521 · cited 1×
“This way wemay vary the level of oversampling according to example types (first idea can be tracked to ADASYN method that ignored safe examples [23]) or supervise the undersampling procedure in order not to discard the important minority representatives.”
MethodThe class imbalance problem: A systematic study1
Japkowicz, N., & Stephen, S. (2002) · Intelligent Data Analysis · doi:10.3233/ida-2002-6504 · Cited by 3,295 · cited 1×
“This domain is known as learning from imbalanced data [29].”
BackgroundGalar, M., Fernandez, A., Barrenechea, E., Bustince, H., & Herrera, F. (2012) · IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews) · doi:10.1109/tsmcc.2011.2161285 · Cited by 2,794 · cited 2×
“[17], an indepth insight into imbalanced data characteristics by López et al.”
Addressing the Curse of Imbalanced Training Sets: One-Sided Selection.
Kubát, M., & Matwin, S. (1997) · Cited by 2,230
CLASSIFICATION OF IMBALANCED DATA: A REVIEW
SUN, Y., WONG, A. K. C., & KAMEL, M. S. (2009) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001409007326 · Cited by 1,692 · cited 2×
“Binary classification problems can be considered as the most developed branch of learning from imbalanced data [51].”
BackgroundLópez, V., Fernández, A., García, S., Palade, V., & Herrera, F. (2013) · Information Sciences · doi:10.1016/j.ins.2013.07.007 · Cited by 1,590 · cited 1×
“[36] and discussion on new perspectives for evaluation classifiers on skewed datasets [42] deserve mentioning.”
BackgroundData Preprocessing in Data Mining
García, S., Luengo, J., & Herrera, F. (2015) · Intelligent Systems Reference Library · doi:10.1007/978-3-319-10247-4 · Cited by 1,368
Vehicle detection in aerial imagery : A small target detection benchmark
Razakarivony, S., & Jurie, F. (2016) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2015.11.002 · Cited by 916 · cited 1×
“Target detection [45] Classification of specified targets appearing with varied frequency (multi-class problem)”
BackgroundA survey of multiple classifier systems as hybrid systems
Woźniak, M., Graña, M., & Corchado, E. (2014) · Information Fusion · doi:10.1016/j.inffus.2013.04.006 · Cited by 912 · cited 1×
“Merging data-level solutions with classifier ensembles [64], resulting in robust and efficient learners [34] is highly popular.”
BackgroundImbalanced Learning: Foundations, Algorithms, and Applications
(2013) · Wiley eBooks · doi:10.1002/9781118646106 · Cited by 831 · cited 1×
“A more recent book in a form of paper collection was edited by He and Ma [25], covering such important issues as sampling strategies, active learning and streaming data.”
BackgroundSáez, J. A., Luengo, J., Stefanowski, J., & Herrera, F. (2015) · Information Sciences · doi:10.1016/j.ins.2014.08.051 · Cited by 627
A novelty detection approach to classification
Japkowicz, N., Myers, C. E., & Gluck, M. A. (1995) · Cited by 412
To Combat Multi-Class Imbalanced Problems by Means of Over-Sampling Techniques
Abdi, L., & Hashemi, S. (2016) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2015.2458858 · Cited by 411 · cited 1×
“Simple re-balancing towards the biggest or smallest class is not a proper approach [1,15].”
BackgroundResampling-Based Ensemble Methods for Online Class Imbalance Learning
Wang, S., Minku, L. L., & Yao, X. (2015) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2014.2345380 · Cited by 396
Fernández, A., López, V., Galar, M., del Jesus, M. J., & Herrera, F. (2013) · Knowledge-Based Systems · doi:10.1016/j.knosys.2013.01.018 · Cited by 388 · cited 3×
“• Decomposition methods used so far [14] apply identical approach for each pair of classes.”
BackgroundMethodInfluentialActive Learning With Drifting Streaming Data
Zliobaite, I., Bifet, A., Pfahringer, B., & Holmes, G. (2014) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2012.2236570 · Cited by 387 · cited 1×
“, by active learning) is currently of crucial importance in this field [69].”
BackgroundMachine Learning from Imbalanced Data Sets 101
Provost, F. J. (2008) · Faculty Digital Archive (New York University Florence) · Cited by 378
Cost-sensitive decision tree ensembles for effective imbalanced classification
Krawczyk, B., Woźniak, M., & Schaefer, G. (2014) · Applied Soft Computing · doi:10.1016/j.asoc.2013.08.014 · Cited by 361 · cited 2×
“Merging data-level solutions with classifier ensembles [64], resulting in robust and efficient learners [34] is highly popular.”
BackgroundSelf-citationON MULTI‐CLASS COST‐SENSITIVE LEARNING
Zhou, Z., & Liu, X. (2010) · Computational Intelligence · doi:10.1111/j.1467-8640.2010.00358.x · Cited by 350 · cited 1×
“The most popular branch is cost-sensitive approaches [67].”
BackgroundManaging Diversity in Regression Ensembles
Brown, G., Wyatt, J., & Peter Tiňo, B. (2005) · Research Explorer (The University of Manchester) · doi:10.5555/1046920.1194899 · Cited by 336
On the use of MapReduce for imbalanced big data using Random Forest
del Río, S., López, V., Benítez, J. M., & Herrera, F. (2014) · Information Sciences · doi:10.1016/j.ins.2014.03.043 · Cited by 328 · cited 1×
“– SMOTE-based oversampling methods applied in distributed environments such as MapReduce tend to fail [13].”
MethodEffective detection of sophisticated online banking fraud on extremely imbalanced data
Wei, W., Li, J., Cao, L., Ou, Y., & Chen, J. (2013) · World Wide Web · doi:10.1007/s11280-012-0178-0 · Cited by 325 · cited 1×
“In real-life applications such as fraud detection [61] or cheminformatics [12] we may deal with problems with imbalance ratio ranging from1:1000 up to 1:5000.”
BackgroundLearning from streaming data with concept drift and imbalance: an overview
Hoens, T. R., Polikar, R., & Chawla, N. V. (2012) · Progress in Artificial Intelligence · doi:10.1007/s13748-011-0008-0 · Cited by 301 · cited 1×
“Dynamical nature of data that arrive either in batches or online poses new challengeswhen imbalanced distributions are to be expected [26].”
BackgroundTypes of minority class examples and their influence on learning classifiers from imbalanced data
Napierala, K., & Stefanowski, J. (2016) · Journal of Intelligent Information Systems · doi:10.1007/s10844-015-0368-1 · Cited by 290 · cited 1×
“Recent work of Napierala and Stefanowski [40] proposed to analyze the neighborhood of each minority class example and assign it to one of four predefined groups: safe, borderline, rare and outliers.”
BackgroundEvolutionary undersampling boosting for imbalanced classification of breast cancer malignancy
Krawczyk, B., Galar, M., Jeleń, Ł., & Herrera, F. (2016) · Applied Soft Computing · doi:10.1016/j.asoc.2015.08.060 · Cited by 263 · cited 1×
“Cancer malignancy grading [30] Analyzing the cancer severity (binary and multi-class problem)”
BackgroundSelf-citationAn improved algorithm for neural network classification of imbalanced training sets
Anand, R., Mehrotra, K. G., Mohan, C. K., & Ranka, S. (1993) · IEEE Transactions on Neural Networks · doi:10.1109/72.286891 · Cited by 262 · cited 1×
“One of the first deeper analyses of learning from skewed data was related to convergence rates of backpropagation-trained neural networks [2].”
BackgroundHellinger distance decision trees are robust and skew-insensitive
Cieslak, D. A., Hoens, T. R., Chawla, N. V., & Kegelmeyer, W. P. (2012) · Data Mining and Knowledge Discovery · doi:10.1007/s10618-011-0222-1 · Cited by 248 · cited 1×
“Recently a Hellinger-distance modification of decision trees [10] and neural networks [66] were proposed and proved to work highly efficiently.”
MLSMOTE: Approaching imbalanced multilabel learning through synthetic instance generation
Charte, F., Rivera, A. J., del Jesus, M. J., & Herrera, F. (2015) · Knowledge-Based Systems · doi:10.1016/j.knosys.2015.07.019 · Cited by 246 · cited 1×
“In multi-label learning so far measures of imbalance and SMOTE-based oversampling approach have been proposed [8].”
Sáez, J. A., Krawczyk, B., & Woźniak, M. (2016) · Pattern Recognition · doi:10.1016/j.patcog.2016.03.012 · Cited by 242 · cited 1×
“difficulty may significantly boost the multi-class performance [46].”
BackgroundSelf-citationNeighbourhood sampling in bagging for imbalanced data
Błaszczyński, J., & Stefanowski, J. (2015) · Neurocomputing · doi:10.1016/j.neucom.2014.07.064 · Cited by 211 · cited 2×
“Preliminary works present in the literature show the high potential of this solution [4,32].”
BackgroundSubspace Learning and Imputation for Streaming Big Data Matrices and Tensors
Mardani, M., Mateos, G., & Giannakis, G. B. (2015) · IEEE Transactions on Signal Processing · doi:10.1109/tsp.2015.2417491 · Cited by 187 · cited 1×
“– We must develop methods for processing and classifying big data in form of graphs, xml structures, video sequences, hyperspectral images, associations, tensors etc [7,37].”
MethodClass imbalance revisited: a new experimental setup to assess the performance of treatment methods
Prati, R. C., Batista, G. E. A. P. A., & Silva, D. F. (2015) · Knowledge and Information Systems · doi:10.1007/s10115-014-0794-3 · Cited by 181 · cited 1×
“[36] and discussion on new perspectives for evaluation classifiers on skewed datasets [42] deserve mentioning.”
BackgroundA dynamic over-sampling procedure based on sensitivity for multi-class problems
Fernández‐Navarro, F., Hervás‐Martínez, C., & Gutiérrez, P. A. (2011) · Pattern Recognition · doi:10.1016/j.patcog.2011.02.019 · Cited by 167 · cited 1×
“Simple re-balancing towards the biggest or smallest class is not a proper approach [1,15].”
BackgroundSiers, M. J., & Islam, M. Z. (2015) · Information Systems · doi:10.1016/j.is.2015.02.006 · Cited by 165 · cited 1×
“Software defect prediction [48] Recognition of errors in code blocks (binary problem)”
BackgroundResampling strategies for regression
Torgo, L., Branco, P., Ribeiro, R. P., & Pfahringer, B. (2015) · Expert Systems · doi:10.1111/exsy.12081 · Cited by 164 · cited 1×
“Main works done on this topic include proposal of evaluation metrics that take into account varying importance of observations [54] and adaptation of undersampling and SMOTE to continuous output prediction problems [53].”
BackgroundA Survey of Predictive Modelling under Imbalanced Distributions
Branco, P., Torgo, L., & Ribeiro, R. P. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1505.01658 · Cited by 162 · cited 2×
“skewed data was proposed by Branco [5] and concentrates”
Triguero, I., del Río, S., López, V., Bacardit, J., Benítez, J. M., & Herrera, F. (2015) · Knowledge-Based Systems · doi:10.1016/j.knosys.2015.05.027 · Cited by 150 · cited 1×
“Big data can also be affected by class imbalance, posing increased challenge to learning systems [55].”
BackgroundDealing with Data Difficulty Factors While Learning from Imbalanced Data
Stefanowski, J. (2015) · Studies in computational intelligence · doi:10.1007/978-3-319-18781-5_17 · Cited by 146 · cited 1×
“This family of algorithms also consists of solutions for cleaning overlapping objects and removing noisy examples that may negatively affect learners [49].”
BackgroundBehavioral Analysis of Insider Threat: A Survey and Bootstrapped Prediction in Imbalanced Data
Azaria, A., Richardson, A., Kraus, S., & Subrahmanian, V. S. (2014) · IEEE Transactions on Computational Social Systems · doi:10.1109/tcss.2014.2377811 · Cited by 133 · cited 1×
“Behavior analysis [3] Recognition of dangerous behavior (binary problem)”
BackgroundPrecision and Recall for Regression
Torgo, L., & Ribeiro, R. P. (2009) · Lecture notes in computer science · doi:10.1007/978-3-642-04747-3_26 · Cited by 123 · cited 1×
“Main works done on this topic include proposal of evaluation metrics that take into account varying importance of observations [54] and adaptation of undersampling and SMOTE to continuous output prediction problems [53].”
BackgroundAdaptive weighted imbalance learning with application to abnormal activity recognition
Gao, X., Chen, Z., Tang, S., Zhang, Y., & Li, J. (2015) · Neurocomputing · doi:10.1016/j.neucom.2015.09.064 · Cited by 112 · cited 1×
“Activity recognition [19] Detection of rare or less-frequent activities (multi-class problem)”
BackgroundWang, S., Li, Z., Chao, W., & Cao, Q. (2012) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2012.6252696 · Cited by 103 · cited 2×
“A class may be a majority one when it is compared to some other classes, but a minority or well-balanced for the rest of them [57].”
BackgroundGalar, M., Fernández, A., Barrenechea, E., Bustince, H., & Herrera, F. (2016) · Information Sciences · doi:10.1016/j.ins.2016.02.056 · Cited by 101 · cited 1×
“Additionally, ensemble pruning techniques dedicated specifically to imbalanced problems should be developed [18].”
BackgroundYu, H., Sun, C., Yang, X., Yang, W., Shen, J., & Qi, Y. (2015) · Knowledge-Based Systems · doi:10.1016/j.knosys.2015.10.012 · Cited by 98 · cited 2×
“Recent studies show that weighting or putting a threshold on continuous output of a classifier (known also as support functions or class probability estimates) can often lead to better results than data resampling andmay be applied to any conventional classifier [31,66].”
BackgroundRamentol, E., Torné, I. G., Choy, S. L., Bello, R., Caballero, Y., Cornelis, C., & Herrera, F. (2015) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2015.10.009 · Cited by 92 · cited 1×
“Industrial systems monitoring [44] Fault detection in industrial machinery (binary problem)”
BackgroundWord Embedding Composition for Data Imbalances in Sentiment and Emotion Classification
Xu, R., Chen, T., Xia, Y., Lu, Q., Liu, B., & Wang, X. (2015) · Cognitive Computation · doi:10.1007/s12559-015-9319-y · Cited by 88 · cited 1×
“Sentiment analysis [65] Emotion and temper recognition in text (binary and multi-class problem)”
BackgroundOn the usefulness of one-class classifier ensembles for decomposition of multi-class problems
Krawczyk, B., Woźniak, M., & Herrera, F. (2015) · Pattern Recognition · doi:10.1016/j.patcog.2015.06.001 · Cited by 85 · cited 2×
“Alternatively, decomposition with one-class classifiers can be considered, as they are robust to class imbalance and can serve as an efficient tool for dealing with difficult multi-class datasets [33].”
BackgroundSelf-citationTabor, J., & Spurek, P. (2014) · Pattern Recognition · doi:10.1016/j.patcog.2014.03.006 · Cited by 82 · cited 1×
“This is especially vivid in case of centroidbased approaches [60], while density-based ones seem to display some robustness to it [52].”
BackgroundTriguero, I., García, S., & Herrera, F. (2014) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2014.2332003 · Cited by 76 · cited 1×
“This phenomenon appears often in semi-supervised [56], active [68] and unsupervised learning [41], especially in clustering.”
BackgroundObject Detection and Recognition in Digital Images: Theory and Practice
Cyganek, B. (2013) · Cited by 64
Object Detection and Recognition in Digital Images
Cyganek, B. (2013) · Wiley eBooks · doi:10.1002/9781118618387 · Cited by 49 · cited 1×
“Big imbalanced data can originate fromvarious specific areas like social networks [27] or computer vision [11] that forces us to work with specific types of data like graphs, tensors or video sequences.”
BackgroundA hybrid decision tree training method using data streams
Woźniak, M. (2010) · Knowledge and Information Systems · doi:10.1007/s10115-010-0345-5 · Cited by 46 · cited 1×
“Mining data streams is one of the important contemporary topics in machine learning [16,62].”
BackgroundBoosted SVM with active learning strategy for imbalanced data
Zięba, M., & Tomczak, J. M. (2014) · Soft Computing · doi:10.1007/s00500-014-1407-5 · Cited by 44 · cited 1×
“This phenomenon appears often in semi-supervised [56], active [68] and unsupervised learning [41], especially in clustering.”
BackgroundData stream mining in ubiquitous environments: state‐of‐the‐art and current directions
Gaber, M. M., da Gama, J. M. P., Krishnaswamy, S. P., Gomes, J. B., & Stahl, F. T. (2014) · Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery · doi:10.1002/widm.1115 · Cited by 44
Gao, Z., Zhang, L., Chen, M., Hauptmann, A. G., Zhang, H., & Cai, A. (2012) · Multimedia Tools and Applications · doi:10.1007/s11042-012-1071-7 · Cited by 43 · cited 1×
“Video mining [20] Recognizing objects and actions in video sequences (binary and multi-class problem)”
BackgroundImbalanced Hyperspectral Image Classification Based on Maximum Margin
Sun, T., Jiao, L., Feng, J., Liu, F., & Zhang, X. (2014) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2014.2349272 · Cited by 40 · cited 1×
“Hyperspectral data analysis [50] Classification of varying areas in multi-dimensional images (multi-class problem)”
BackgroundSelf-training in significance space of support vectors for imbalanced biomedical event data
Munkhdalai, T., Namsrai, O.-E., & Ryu, K. H. (2015) · BMC Bioinformatics · doi:10.1186/1471-2105-16-s7-s6 · Cited by 37 · cited 1×
“Text mining [39] Detecting relations in literature (binary problem)”
BackgroundWeighted one-class classification for different types of minority class examples in imbalanced data
Krawczyk, B., Woźniak, M., & Herrera, F. (2014) · IEEE Symposium on Computational Intelligence and Data Mining (CIDM) · doi:10.1109/cidm.2014.7008687 · Cited by 35 · cited 1×
“Preliminary works present in the literature show the high potential of this solution [4,32].”
BackgroundSelf-citationOnline cost-sensitive neural network classifiers for non-stationary and imbalanced data streams
Ghazikhani, A., Monsefi, R., & Yazdi, H. S. (2012) · Neural Computing and Applications · doi:10.1007/s00521-012-1071-6 · Cited by 34 · cited 1×
“in recent yearswith singlemodel approaches [22] and ensemble learning [58] there are still numerous open challenges to be addressed.”
BackgroundAn unsupervised self-organizing learning with support vector ranking for imbalanced datasets
Nguwi, Y.-Y., & Cho, S. (2010) · Expert Systems with Applications · doi:10.1016/j.eswa.2010.05.054 · Cited by 32
Cost-Sensitive Neural Network with ROC-Based Moving Threshold for Imbalanced Classification
Krawczyk, B., & Woźniak, M. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-24834-9_6 · Cited by 32 · cited 1×
“Recent studies show that weighting or putting a threshold on continuous output of a classifier (known also as support functions or class probability estimates) can often lead to better results than data resampling andmay be applied to any conventional classifier [31,66].”
BackgroundSelf-citationResampling and Cost-Sensitive Methods for Imbalanced Multi-instance Learning
Wang, X., Liu, X., Japkowicz, N., & Matwin, S. (2013) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw.2013.85 · Cited by 18 · cited 1×
“In multi-instance learning cost-sensitive and standard resampling have been used to counter skewed distributions with regard to number of bags and number of instanceswithin bags [38,59].”
MethodCzarnecki, W. M., & Rataj, K. (2015) · IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom) · doi:10.1109/trustcom.2015.581 · Cited by 13 · cited 1×
“In real-life applications such as fraud detection [61] or cheminformatics [12] we may deal with problems with imbalance ratio ranging from1:1000 up to 1:5000.”
BackgroundStructural XML Classification in Concept Drifting Data Streams
Brzeziński, D., & Piernik, M. (2015) · New Generation Computing · doi:10.1007/s00354-015-0401-5 · Cited by 10 · cited 1×
“– We must develop methods for processing and classifying big data in form of graphs, xml structures, video sequences, hyperspectral images, associations, tensors etc [7,37].”
MethodMulti-exemplar based clustering for imbalanced data
Wang, Y., & Chen, L. (2014) · International Conference on Control, Automation, Robotics and Vision (ICARCV) · doi:10.1109/icarcv.2014.7064454 · Cited by 9 · cited 1×
“This is especially vivid in case of centroidbased approaches [60], while density-based ones seem to display some robustness to it [52].”
BackgroundA Bag Oversampling Approach for Class Imbalance in Multiple Instance Learning
Mera, C., Arrieta, J., Orozco‐Alzate, M., & Branch, J. W. (2015) · Lecture notes in computer science · doi:10.1007/978-3-319-25751-8_87 · Cited by 9 · cited 1×
“In multi-instance learning cost-sensitive and standard resampling have been used to counter skewed distributions with regard to number of bags and number of instanceswithin bags [38,59].”
MethodWoźniak, M. (2013) · Studies in computational intelligence · doi:10.1007/978-3-642-40997-4_3 · Cited by 6
A Classifier Ensembling Approach for Imbalanced Social Link Prediction
Hurtado, J., Taweewitchakreeya, N., Kong, X., & Xingquan Zhu, A. (2013) · International Conference on Machine Learning and Applications · doi:10.1109/icmla.2013.88 · Cited by 4 · cited 1×
“Big imbalanced data can originate fromvarious specific areas like social networks [27] or computer vision [11] that forces us to work with specific types of data like graphs, tensors or video sequences.”
Background
Cited by
Survey on deep learning with class imbalance
Johnson, J. M., & Khoshgoftaar, T. M. (2019) · Journal of Big Data · doi:10.1186/s40537-019-0192-5 · Cited by 3,032
SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary
Fernandez, A., Garcia, S., Herrera, F., & Chawla, N. V. (2018) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.11192 · Cited by 2,219
“…networks, and medical diagnosis), providing a novel and contemporaneous set of challenges to the machine learning and data science researchers (Krawczyk, 2016; Haixiang et al., 2017; Maua & Galinac Grbac, 2017; Zhang et al., 2017; Zuo et al., 2016; Lichtenwalter et al., 2010; Krawczyk et al.,…”
BackgroundLuque, A., Carrasco, A., Martín, A., & de las Heras, A. (2019) · Pattern Recognition · doi:10.1016/j.patcog.2019.02.023 · Cited by 1,128
Ensemble learning for data stream analysis: A survey
Krawczyk, B., Minku, L. L., Gama, J., Stefanowski, J., & Woźniak, M. (2017) · Information Fusion · doi:10.1016/j.inffus.2017.02.004 · Cited by 1,050
“It has been more studied in this static framework and many new algorithms have already been introduced, for their comprehensive review see the recent monograph [73] or surveys [72,101,172] .”
Self-citationHyperparameter optimization: Foundations, algorithms, best practices, and open challenges
Bischl, B., Binder, M., Lang, M., Pielok, T., Richter, J., Coors, S., Thomas, J., Ullmann, T., Becker, M., Boulesteix, A., et al. (2023) · WIREs Data Mining and Knowledge Discovery · doi:10.1002/widm.1484 · Cited by 930
Metadata
- Authors
- Krawczyk, B.
- Affiliations
- Wrocław University of Science and Technology (PL)
- Corresponding author
- Krawczyk, Bartosz
- ORCID
- Krawczyk, Bartosz
- Venue
- Progress in Artificial Intelligence
- Published
- 2016-04-22
- Volume
- 5
- Issue
- 4
- Pages
- 221-232
- ISSN
- 2192-6352 (ISSN-L)
- 2192-6360
- OpenAlex
- W2338318698
- Cited by
- 2,459
- References
- 70
- Citations by year
- 2026: 233 · 2025: 299 · 2024: 340 · 2023: 337 · 2022: 340 · 2021: 343 · 2020: 240 · 2019: 167 · 2018: 104 · 2017: 53 · 2016: 2
- FWCI
- 89.32
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Locations
- Progress in Artificial Intelligence (publishedVersion, cc-by)
- Funders
- Polish National Science Center (DEC-2013/09/B/ST6/02264)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Imbalanced Data Classification Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Machine Learning and Data Classification
- Physical Sciences › Computer Science › Artificial Intelligence › Data Stream Mining Techniques
- Keywords
- imbalanced learning, class imbalance, skewed distributions, binary classification, regression analysis, clustering, data streams, big data analytics, social media applications, computer vision applications, real-time learning, adaptive learning, computational efficiency
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.