Pub Date : 2023-09-12DOI: 10.1007/s41060-023-00445-1
Mythreyi Velmurugan, Chun Ouyang, Renuka Sindhgatta, Catarina Moreira
Abstract Modern machine learning methods allow for complex and in-depth analytics, but the predictive models generated by these methods are often highly complex and lack transparency. Explainable Artificial Intelligence (XAI) methods are used to improve the interpretability of these complex “black box” models, thereby increasing transparency and enabling informed decision-making. However, the inherent fitness of these explainable methods, particularly the faithfulness of explanations to the decision-making processes of the model, can be hard to evaluate. In this work, we examine and evaluate the explanations provided by four XAI methods, using fully transparent “glass box” models trained on tabular data. Our results suggest that the fidelity of explanations is determined by the types of variables used, as well as the linearity of the relationship between variables and model prediction. We find that each XAI method evaluated has its own strengths and weaknesses, determined by the assumptions inherent in the explanation mechanism. Thus, though such methods are model-agnostic, we find significant differences in explanation quality across different technical setups. Given the numerous factors that determine the quality of explanations, including the specific explanation-generation procedures implemented by XAI methods, we suggest that model-agnostic XAI methods may still require expert guidance for implementation.
{"title":"Through the looking glass: evaluating post hoc explanations using transparent models","authors":"Mythreyi Velmurugan, Chun Ouyang, Renuka Sindhgatta, Catarina Moreira","doi":"10.1007/s41060-023-00445-1","DOIUrl":"https://doi.org/10.1007/s41060-023-00445-1","url":null,"abstract":"Abstract Modern machine learning methods allow for complex and in-depth analytics, but the predictive models generated by these methods are often highly complex and lack transparency. Explainable Artificial Intelligence (XAI) methods are used to improve the interpretability of these complex “black box” models, thereby increasing transparency and enabling informed decision-making. However, the inherent fitness of these explainable methods, particularly the faithfulness of explanations to the decision-making processes of the model, can be hard to evaluate. In this work, we examine and evaluate the explanations provided by four XAI methods, using fully transparent “glass box” models trained on tabular data. Our results suggest that the fidelity of explanations is determined by the types of variables used, as well as the linearity of the relationship between variables and model prediction. We find that each XAI method evaluated has its own strengths and weaknesses, determined by the assumptions inherent in the explanation mechanism. Thus, though such methods are model-agnostic, we find significant differences in explanation quality across different technical setups. Given the numerous factors that determine the quality of explanations, including the specific explanation-generation procedures implemented by XAI methods, we suggest that model-agnostic XAI methods may still require expert guidance for implementation.","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135878622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-09DOI: 10.1007/s41060-023-00444-2
Chisimkwuo John, Emmanuel J. Ekpenyong, Charles Chinedu Nworu, Chukwuemeka O. Omekara
{"title":"A new robust bootstrapped singular value decomposition algorithm using the sample myriad estimate","authors":"Chisimkwuo John, Emmanuel J. Ekpenyong, Charles Chinedu Nworu, Chukwuemeka O. Omekara","doi":"10.1007/s41060-023-00444-2","DOIUrl":"https://doi.org/10.1007/s41060-023-00444-2","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136192381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-07DOI: 10.1007/s41060-023-00440-6
Jurgen van den Hoogen, Dan Hudson, Stefan Bloemheuvel, Martin Atzmueller
Abstract Industrial fault detection has become more data-driven due to advancements in automated data analysis using deep learning. Such methods make it possible to extract useful features, e. g., from time series data retrieved from sensors, which is typically of complex nature. This allows for effective fault detection and prognostics that boost the efficiency and productivity of industrial equipment. This work explores the influence of a variety of architectural hyperparameters on the performance of one-dimensional convolutional neural networks (CNN). Using a multi-method approach, this paper focuses specifically on wide-kernel CNN models for industrial fault detection, that have proven to perform well for tasks such as classifying vibration signals retrieved from sensors. By varying hyperparameters such as the kernel size, stride and number of filters, an extensive hyperparameter space search was conducted; to identify optimal settings, we collected a total of 12,960 different combinations on three datasets into a model hyperparameter dataset, with their respective performance on the underlying fault detection task. Afterwards, this dataset was explored with follow-up analysis including statistical, feature, pattern and hyperparameter impact analysis. We find that although performance varies substantially depending on hyperparameter choices, there is no single simple strategy to optimise performance across the three datasets. However, an optimal setting in terms of performance can be found in the number of filters used in the later layers of the architecture for all datasets. Furthermore, hyperparameter importance differs across and within the datasets, and we found nonlinear relationships between hyperparameter settings and performance. Our analysis highlights key considerations when applying a wide-kernel CNN architecture to new data within the field of industrial fault detection. This supports practitioners who wish to apply and train state-of-the-art convolutional learning methods to apply to similar fault detection settings, e. g., vibration data arising from new combinations of sensors and/or machinery in the context of bearing faults.
{"title":"Hyperparameter analysis of wide-kernel CNN architectures in industrial fault detection: an exploratory study","authors":"Jurgen van den Hoogen, Dan Hudson, Stefan Bloemheuvel, Martin Atzmueller","doi":"10.1007/s41060-023-00440-6","DOIUrl":"https://doi.org/10.1007/s41060-023-00440-6","url":null,"abstract":"Abstract Industrial fault detection has become more data-driven due to advancements in automated data analysis using deep learning. Such methods make it possible to extract useful features, e. g., from time series data retrieved from sensors, which is typically of complex nature. This allows for effective fault detection and prognostics that boost the efficiency and productivity of industrial equipment. This work explores the influence of a variety of architectural hyperparameters on the performance of one-dimensional convolutional neural networks (CNN). Using a multi-method approach, this paper focuses specifically on wide-kernel CNN models for industrial fault detection, that have proven to perform well for tasks such as classifying vibration signals retrieved from sensors. By varying hyperparameters such as the kernel size, stride and number of filters, an extensive hyperparameter space search was conducted; to identify optimal settings, we collected a total of 12,960 different combinations on three datasets into a model hyperparameter dataset, with their respective performance on the underlying fault detection task. Afterwards, this dataset was explored with follow-up analysis including statistical, feature, pattern and hyperparameter impact analysis. We find that although performance varies substantially depending on hyperparameter choices, there is no single simple strategy to optimise performance across the three datasets. However, an optimal setting in terms of performance can be found in the number of filters used in the later layers of the architecture for all datasets. Furthermore, hyperparameter importance differs across and within the datasets, and we found nonlinear relationships between hyperparameter settings and performance. Our analysis highlights key considerations when applying a wide-kernel CNN architecture to new data within the field of industrial fault detection. This supports practitioners who wish to apply and train state-of-the-art convolutional learning methods to apply to similar fault detection settings, e. g., vibration data arising from new combinations of sensors and/or machinery in the context of bearing faults.","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135046940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-07DOI: 10.1007/s41060-023-00446-0
Yuru Teng, Guotao Wang, Cailing He, Yaoyang Wu, Chaoran Li
{"title":"Optimization of Dynamic Time Warping Algorithm for Abnormal Signal Detection","authors":"Yuru Teng, Guotao Wang, Cailing He, Yaoyang Wu, Chaoran Li","doi":"10.1007/s41060-023-00446-0","DOIUrl":"https://doi.org/10.1007/s41060-023-00446-0","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"247 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2023-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74060407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-26DOI: 10.1007/s41060-023-00447-z
A. Barrick, I. Métais, H. Ettajani, J. Marion, A. Châtel
{"title":"Establishing FAIR (Findable, Accessible, Interoperable and Reusable) principles for estuarine organisms exposed to engineered nanomaterials","authors":"A. Barrick, I. Métais, H. Ettajani, J. Marion, A. Châtel","doi":"10.1007/s41060-023-00447-z","DOIUrl":"https://doi.org/10.1007/s41060-023-00447-z","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"220 1","pages":"407 - 419"},"PeriodicalIF":2.4,"publicationDate":"2023-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79820342","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-25DOI: 10.1007/s41060-023-00442-4
Marko Harasic, Felix-Sebastian Keese, Denny Mattern, A. Paschke
{"title":"Recent advances and future challenges in federated recommender systems","authors":"Marko Harasic, Felix-Sebastian Keese, Denny Mattern, A. Paschke","doi":"10.1007/s41060-023-00442-4","DOIUrl":"https://doi.org/10.1007/s41060-023-00442-4","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"39 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2023-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91137197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-19DOI: 10.1007/s41060-023-00435-3
D. Kanzari
{"title":"Context-adaptive intelligent agents behaviors: multivariate LSTM-based decision making on the cryptocurrency market","authors":"D. Kanzari","doi":"10.1007/s41060-023-00435-3","DOIUrl":"https://doi.org/10.1007/s41060-023-00435-3","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"48 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2023-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75545569","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"CIAMS: clustering indices-based automatic classification model selection","authors":"Sudarsun Santhiappan, Nitin Shravan, Balaraman Ravindran","doi":"10.1007/s41060-023-00441-5","DOIUrl":"https://doi.org/10.1007/s41060-023-00441-5","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135937636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-16DOI: 10.1007/s41060-023-00433-5
Marco Bombieri, M. Rospocher, Simone Paolo Ponzetto, P. Fiorini
{"title":"Surgicberta: a pre-trained language model for procedural surgical language","authors":"Marco Bombieri, M. Rospocher, Simone Paolo Ponzetto, P. Fiorini","doi":"10.1007/s41060-023-00433-5","DOIUrl":"https://doi.org/10.1007/s41060-023-00433-5","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"31 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2023-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90769418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-16DOI: 10.1007/s41060-023-00437-1
R. Raut, Francesca Spezzano
{"title":"Enhancing hate speech detection with user characteristics","authors":"R. Raut, Francesca Spezzano","doi":"10.1007/s41060-023-00437-1","DOIUrl":"https://doi.org/10.1007/s41060-023-00437-1","url":null,"abstract":"","PeriodicalId":45667,"journal":{"name":"International Journal of Data Science and Analytics","volume":"5 1","pages":""},"PeriodicalIF":2.4,"publicationDate":"2023-08-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81437314","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：