Article
Computer Science, Theory & Methods
Padraig Cunningham, Sarah Jane Delany
Summary: The article provides an overview of Nearest Neighbour classification techniques, focusing on similarity assessment mechanisms, computational issues in identifying nearest neighbours, and methods for reducing the dimension of the data. New sections on similarity measures for time-series, retrieval speedup, and intrinsic dimensionality have been added, along with an Appendix containing Python code for key methods.
ACM COMPUTING SURVEYS
(2021)
Article
Automation & Control Systems
Jiawei Niu, Zhunga Liu, Yao Lu, Zaidao Wen
Summary: The paper introduces a new method called evidential combination of classifiers (ECC) for handling imbalanced data by combining hybrid-sampling, over-sampling, and under-sampling methods. By revising the classification results at the decision level, the method aims to reduce error risk and achieve improved classification performance through evidence theory combination.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Computer Science, Information Systems
Hadi A. Khorshidi, Uwe Aickelin
Summary: The study proposes a new approach using diversity optimization for synthetic instance generation to address imbalanced data in classification. The proposed formulations show competitive performance in improving classifier performance, with DIWO outperforming other comparable methods. Both formulations exhibit robustness by reducing classifier variance.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Artificial Intelligence
Chen Wang, Chengyuan Deng, Zhoulu Yu, Dafeng Hui, Xiaofeng Gong, Ruisen Luo
Summary: The study introduces a novel dynamic ensemble method AER to address the overfitting issue in binary imbalanced data classification through regularization and utilizing global geometry of data, demonstrating superior performance in experiments.
INFORMATION FUSION
(2021)
Article
Engineering, Electrical & Electronic
Natan Kruglyak, Robert Forchheimer
Summary: The paper suggests combining traditional classification methods with artificial neural networks (ANNs) to leverage the benefits of both fields and improve classification performance. By illustrating the approach with a specific example and showing how traditional methods can be translated into ANNs and improved for stability, the study demonstrates the effectiveness of the two-step design philosophy.
INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Pu Zhang, Changjing Shang, Qiang Shen
Summary: This article presents a novel approach for computing interpolated outcomes with TSK models, which improves computational efficiency and minimizes the adverse impact on accuracy. It introduces a rule-clustering-based method for selecting rules for large rule bases. Experimental results show the effectiveness of the introduced techniques.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Remote Sensing
Nathalie Guimaraes, Luis Padua, Joaquim J. Sousa, Albino Bento, Pedro Couto
Summary: In Portugal, almonds are important due to their nutritional properties. This study explores the classification of almond cultivars using remote-sensing data and machine learning classifiers. The results demonstrate the importance of feature selection in optimizing classifier performance.
INTERNATIONAL JOURNAL OF REMOTE SENSING
(2023)
Article
Computer Science, Information Systems
Marek Sikora, Pawel Matyszok, Lukasz Wrobel
Summary: This article introduces an action rule induction Algorithm based on sequential covering, with two variants presented, allowing induction of action rules from a source and target decision class perspective. A recommendation induction method is also presented, based on a set of induced action rules, to recommend actions needed to move examples from a given source class to the appropriate target class.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Yanela Rodriguez Alvarez, Maria Matilde Garcia Lorenzo, Yaile Caballero Mota, Yaima Filiberto Cabrera, Isabel M. Garcia Hilarion, Daniela Machado Montes de Oca, Rafael Bello Perez
Summary: The paper introduces two prototype selection classifiers based on fuzzy logic, demonstrating their good performance and efficiency in dealing with class-imbalanced data.
PATTERN RECOGNITION LETTERS
(2022)
Article
Computer Science, Information Systems
Xiangrui Chao, Gang Kou, Yi Peng, Alberto Fernandez
Summary: Balancing the accuracy rates of majority and minority classes is challenging in imbalanced classification. This study introduces a new criterion, an efficiency curve, to comprehensively evaluate imbalanced classifiers and analyzes the impact of imbalanced ratio and data characteristics on classifier efficiency.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Kyoungok Kim
Summary: This study introduces an enhanced kNN algorithm NCC-kNN for classification of imbalanced datasets, especially performing well on data with low positive class coherence.
PATTERN RECOGNITION
(2021)
Article
Computer Science, Artificial Intelligence
Zhan Ao Huang, Yongsheng Sang, Yanan Sun, Jiancheng Lv
Summary: Most real-life data suffer from imbalance problems, causing negative class preference behavior in neural networks. To address this issue, a new paradigm is proposed, which includes an informative undersampling strategy to solve the problem of gradient inundation and a boundary expansion strategy to alleviate the problem of insufficient empirical representation of positive samples. Experimental results show that the proposed paradigm outperforms existing methods in terms of AUC on multiple imbalanced datasets.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Nawal Almutairi, Frans Coenen, Keith Dures
Summary: Secure collaborative data clustering using SecureCL is based on phi-data implemented with Super Secure Chain Distance Matrices and encrypted with Multi-User Order Preserving Encryption. Unlike other systems, SecureCL does not require user participation or recourse to Secure Multi-Party Computation protocols or 'secret sharing' mechanisms. Experimental results demonstrate that SecureCL can produce securely cluster configurations comparable to those produced using standard, non-encrypted approaches.
Article
Computer Science, Artificial Intelligence
Yang-Geng Fu, Hong-Yun Huang, Yu Guan, Ying-Ming Wang, Wenxi Liu, Wei-Jie Fang
Summary: This study proposes a Belief Rule-Based (BRB) reasoning model based on the evidential reasoning algorithm to address the issue of data imbalance in rule-based systems. By incorporating data-driven characteristics and data class rebalancing techniques, the model dynamically adjusts rule weights to improve classification performance by expanding the proportion of minority classes and building an ensemble classifier with excellent performance.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Medicine, General & Internal
Madhumita Pal, Smita Parija, Ganapati Panda, Kuldeep Dhama, Ranjan K. Mohapatra
Summary: This study utilizes reliable machine learning techniques to achieve automatic detection of cardiovascular disease, achieving high accuracy and performance by processing and optimizing publicly available data. The proposed methodology has the potential for application in other diseases as well.
Article
Computer Science, Artificial Intelligence
Mateusz Lango, Jerzy Stefanowski
JOURNAL OF INTELLIGENT INFORMATION SYSTEMS
(2018)
Article
Computer Science, Information Systems
Dariusz Brzezinski, Jerzy Stefanowski, Robert Susmaga, Izabela Szczech
INFORMATION SCIENCES
(2018)
Article
Computer Science, Artificial Intelligence
Dariusz Brzezinski, Leandro L. Minku, Tomasz Pewinski, Jerzy Stefanowski, Artur Szumaczuk
Summary: Class imbalance poses additional challenges when learning classifiers from concept drifting data streams. Existing work primarily focuses on addressing global imbalance ratio, while neglecting other data complexities. Independent research on static imbalanced data has emphasized the influential role of local data difficulty factors. Investigating the interactions between concept drifts and local data difficulty factors in concept drifting data streams is crucial, as revealed by our comprehensive study.
KNOWLEDGE AND INFORMATION SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Mateusz Lango, Jerzy Stefanowski
Summary: This study experimentally investigates the impact of various multi-class imbalanced difficulty factors on the performance of classifiers. The results reveal that class overlapping and class size configurations are important difficulties.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Witold Andrzejewski, Jedrzej Potoniec, Maciej Drozdowski, Jerzy Stefanowski, Robert Wrembel, Pawel Stapf
Summary: Effective heat energy demand prediction is crucial in combined heat power systems. Existing algorithms do not adequately consider computational costs and ease of implementation in industrial systems. This paper proposes two types of algorithms for heat demand prediction and evaluates them experimentally in terms of prediction quality and computational cost.
DATABASE AND EXPERT SYSTEMS APPLICATIONS, DEXA 2022, PT I
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Kamil Plucinski, Mateusz Lango, Jerzy Stefanowski
Summary: This paper introduces a new prototype-based convolutional neural architecture for text classification, which offers faithful predictions' explanations compared to traditional attention mechanisms. It also demonstrates that dynamic tuning of the number of prototypes can lead to performance gains.
MACHINE LEARNING AND PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021, PT I
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Jacek Grycza, Damian Horna, Hanna Klimczak, Mateusz Lango, Kamil Plucinski, Jerzy Stefanowski
Summary: Multi-imbalance is an open-source Python library designed to provide the Python community with tools for handling multi-class imbalanced problems. It includes implementations of binary decomposition techniques, ensembles, and a variety of re-sampling approaches for multi-class imbalanced classification.
MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES: APPLIED DATA SCIENCE AND DEMO TRACK, ECML PKDD 2020, PT V
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Piotr Janiszewski, Mateusz Lango, Jerzy Stefanowski
Summary: Online harassment is a significant issue in modern societies, often mitigated by the manual work of website moderators and supported by machine learning tools. Previous methods only allow for retrospective detection of online abuse, while proactive approaches have been proposed to help moderators prevent conversation breakdown. This study introduces a new method based on deep neural networks that predicts the likelihood of conversation breakdown and the time remaining until derailment, showing improvement over current state-of-the-art methods.
MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2021: APPLIED DATA SCIENCE TRACK, PT V
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Jerzy Stefanowski
Summary: This paper summarizes the difficulty factors and research status of multiple class imbalanced problem, and presents three new methods for learning classifiers from multi-class imbalanced data.
ROUGH SETS (IJCRS 2021)
(2021)
Article
Computer Science, Artificial Intelligence
Grzegorz J. Nalepa, Jerzy Stefanowski
FOUNDATIONS OF COMPUTING AND DECISION SCIENCES
(2020)
Article
Computer Science, Artificial Intelligence
Dariusz Brzezinski, Jerzy Stefanowski, Robert Susmaga, Izabela Szczech
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2020)
Article
Automation & Control Systems
Malgorzata Janicka, Mateusz Lango, Jerzy Stefanowski
INTERNATIONAL JOURNAL OF APPLIED MATHEMATICS AND COMPUTER SCIENCE
(2019)
Proceedings Paper
Computer Science, Artificial Intelligence
Szymon Wojciechowski, Szymon Wilk, Jerzy Stefanowski
PROCEEDINGS OF THE 10TH INTERNATIONAL CONFERENCE ON COMPUTER RECOGNITION SYSTEMS CORES 2017
(2018)
Proceedings Paper
Computer Science, Artificial Intelligence
Dariusz Brzezinski, Jerzy Stefanowski, Robert Susmaga, Izabela Szczech
MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2017, PT III
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Mateusz Lango, Dariusz Brzezinski, Sebastian Firlik, Jerzy Stefanowski
DISCOVERY SCIENCE, DS 2017
(2017)