Data quality measures based on granular computing for multi-label classification

Bello, Marilyn; Nápoles, Gonzalo; Vanhoof, Koen; Bello, Rafael

doi:10.1016/j.ins.2021.01.027

Cited by 23 publications

(6 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This concern is magnified in multi-label classification, where instances can be associated with multiple labels simultaneously. In this context, AI system robustness is at risk due to diverse types of datalabel issues, although the task can reflect the complex relationships present in real-world data (Bello et al 2021).…”

Section: Introductionmentioning

confidence: 99%

Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise

Song,

Kim,

Lee

2024

AAAI

View full text Add to dashboard Cite

Multi-label classification poses challenges due to imbalanced and noisy labels in training data. In this paper, we propose a unified data augmentation method, named BalanceMix, to address these challenges. Our approach includes two samplers for imbalanced labels, generating minority-augmented instances with high diversity. It also refines multi-labels at the label-wise granularity, categorizing noisy labels as clean, re-labeled, or ambiguous for robust optimization. Extensive experiments on three benchmark datasets demonstrate that BalanceMix outperforms existing state-of-the-art methods. We release the code at https://github.com/DISL-Lab/BalanceMix.

show abstract

Section: Introductionmentioning

confidence: 99%

Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise

Song,

Kim,

Lee

2024

AAAI

View full text Add to dashboard Cite

show abstract

“…Traditional classification algorithms often assume that the input data is crisp and welldefined (Bezdek & Keller, 2020) (Bello et al, 2021). In real-world scenarios, datasets are often characterized by uncertainty, vagueness, and imprecision (Díaz-Rodríguez et al, 2014).…”

Section: Introductionmentioning

confidence: 99%

Integrating hybrid grid partition and rough set method for fuzzy rule generation: a novel approach for accurate dataset classification

Joseph¹

2023

emod

View full text Add to dashboard Cite

Accurate dataset classification is a critical task in various domains, and combining different methodologies can enhance classification performance. This research presents a novel approach that integrates Hybrid Grid Partition and Rough Set methods for fuzzy rule generation, aiming to improve accuracy and interpretability in dataset classification. The proposed approach leverages Hybrid Grid Partition to discretize continuous attributes and Rough Set attribute reduction to identify essential attributes, enabling accurate classification while handling uncertainty and imprecision. The generated fuzzy rules provide interpretability, aiding decision-making processes and providing insights into classification factors. The approach's robustness and generalization capabilities are demonstrated through experiments on diverse datasets, indicating its potential applicability in real-world scenarios. However, limitations such as the absence of specific evaluation metrics and the need for further validation on larger datasets are acknowledged. Overall, this research contributes to accurate dataset classification by offering a novel integrated approach and highlighting areas for future investigation and refinement

show abstract

“…DQ is critical for tracking data value and relevance, and we believe its use in quantifying data will give us a handle on what data are available, what the data’s value might be for business decision making, and whether the data should be assessed primarily during the data transformations at the pre-processing and processing stages of the data. The accuracy of a classification model is heavily dependent on the DQ, so measuring DQ [ 6 ] is critical for estimating task complexity earlier. DQ attributes should be verified, improved, and regulated throughout their life cycle, as they have a direct impact on the conclusions drawn from data analysis.…”

Section: Introductionmentioning

confidence: 99%

Empowering Patient Similarity Networks through Innovative Data-Quality-Aware Federated Profiling

Navaz

Serhani

Kassabi

et al. 2023

Sensors

View full text Add to dashboard Cite

Continuous monitoring of patients involves collecting and analyzing sensory data from a multitude of sources. To overcome communication overhead, ensure data privacy and security, reduce data loss, and maintain efficient resource usage, the processing and analytics are moved close to where the data are located (e.g., the edge). However, data quality (DQ) can be degraded because of imprecise or malfunctioning sensors, dynamic changes in the environment, transmission failures, or delays. Therefore, it is crucial to keep an eye on data quality and spot problems as quickly as possible, so that they do not mislead clinical judgments and lead to the wrong course of action. In this article, a novel approach called federated data quality profiling (FDQP) is proposed to assess the quality of the data at the edge. FDQP is inspired by federated learning (FL) and serves as a condensed document or a guide for node data quality assurance. The FDQP formal model is developed to capture the quality dimensions specified in the data quality profile (DQP). The proposed approach uses federated feature selection to improve classifier precision and rank features based on criteria such as feature value, outlier percentage, and missing data percentage. Extensive experimentation using a fetal dataset split into different edge nodes and a set of scenarios were carefully chosen to evaluate the proposed FDQP model. The results of the experiments demonstrated that the proposed FDQP approach positively improved the DQ, and thus, impacted the accuracy of the federated patient similarity network (FPSN)-based machine learning models. The proposed data-quality-aware federated PSN architecture leveraging FDQP model with data collected from edge nodes can effectively improve the data quality and accuracy of the federated patient similarity network (FPSN)-based machine learning models. Our profiling algorithm used lightweight profile exchange instead of full data processing at the edge, which resulted in optimal data quality achievement, thus improving efficiency. Overall, FDQP is an effective method for assessing data quality in the edge computing environment, and we believe that the proposed approach can be applied to other scenarios beyond patient monitoring.

show abstract

Data quality measures based on granular computing for multi-label classification

Cited by 23 publications

References 27 publications

Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise

Toward Robustness in Multi-Label Classification: A Data Augmentation Strategy against Imbalance and Noise

Integrating hybrid grid partition and rough set method for fuzzy rule generation: a novel approach for accurate dataset classification

Empowering Patient Similarity Networks through Innovative Data-Quality-Aware Federated Profiling

Contact Info

Product

Resources

About