Pub Date : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494409
Nasibeh Rady Raz, M. Akbarzadeh-T.
Social things face great uncertainty and complexity in their decision-making. This is true whether the social thing is as large as the electric grid of a country or as small as drug carrier targeted nanomachines employed for cancer treatment. The problem is further complicated when it is tasked with serving humans, with the intricate and ill-defined meaning of service. Therefore, there is a need for cognitive decision-making in which human factors of time, attitude, attention, trust, and bias are considered in the recommendation, prediction, analysis, estimation, and automated decision making. Here, we introduce a trust-based decision-making architecture for swarms of bioinspired nanomachines. Particularly, the factor of trust is used as an index for swarm joining and disjoining. Each nanomachine's decision is considered as a new attitude that is weakened or reinforced by a trust factor. The trust factor is derived using a Fuzzy Cognitive Map which is composed of integrity, competence, consistency, loyalty, and openness. Nanomachines with high trust factors form a dense group and change to a “trustee” swarm. The trustee converges to the cancer site. The result shows the proposed method in targeted drug delivery outperforms the competing strategies with lower hypoxic and endothelial cell density as the marker of cancer.
{"title":"Trust-based Cognitive Decision Making by Social Things - A Case Study of Cancer Treatment","authors":"Nasibeh Rady Raz, M. Akbarzadeh-T.","doi":"10.1109/FUZZ45933.2021.9494409","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494409","url":null,"abstract":"Social things face great uncertainty and complexity in their decision-making. This is true whether the social thing is as large as the electric grid of a country or as small as drug carrier targeted nanomachines employed for cancer treatment. The problem is further complicated when it is tasked with serving humans, with the intricate and ill-defined meaning of service. Therefore, there is a need for cognitive decision-making in which human factors of time, attitude, attention, trust, and bias are considered in the recommendation, prediction, analysis, estimation, and automated decision making. Here, we introduce a trust-based decision-making architecture for swarms of bioinspired nanomachines. Particularly, the factor of trust is used as an index for swarm joining and disjoining. Each nanomachine's decision is considered as a new attitude that is weakened or reinforced by a trust factor. The trust factor is derived using a Fuzzy Cognitive Map which is composed of integrity, competence, consistency, loyalty, and openness. Nanomachines with high trust factors form a dense group and change to a “trustee” swarm. The trustee converges to the cancer site. The result shows the proposed method in targeted drug delivery outperforms the competing strategies with lower hypoxic and endothelial cell density as the marker of cancer.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"144 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129539428","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494507
Zack Ellerby, Christian Wagner
Capturing interval-valued, as opposed to more conventional point-valued data, offers a potentially efficient method of obtaining richer information in individual responses. In turn, interval-valued data provide a strong foundation for subsequent fuzzy set based modelling-e.g., using the Interval Agreement Approach. In 2019, open-source software (DECSYS) was released to enable digital administration of interval-valued surveys using an ellipse response mode. This study follows on from an appraisal of this software and demonstration of practical value of the approach, reported last year, in one of many potential real-world applications (consumer preference research). A key ambition of ellipse-based interval elicitation is to maximise response efficiency-i.e., minimising workload and complexity in obtaining this richer information. User experience is therefore a vital consideration regarding potential for broader adoption. The present paper documents a direct empirical comparison between interval-valued response elicitation (using ellipses) and a conventional point-valued counterpart (using a Visual Analogue Scale), in terms of user experience during completion of a simple quantitative estimation task. We examine differences in perceived ease-of-use, unnecessary complexity and effective communication of desired responses, as well as overall liking-with positive outcomes for the interval-valued response mode in each case. We also report results of multiple regression analyses examining how the first three variables contribute to participants' overall liking of each response mode, as well as exploring differences driven by potentially important demographic factors (i.e., gender, age & native English speaking).
{"title":"Do People Prefer to Give Interval-Valued or Point Estimates and Why?","authors":"Zack Ellerby, Christian Wagner","doi":"10.1109/FUZZ45933.2021.9494507","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494507","url":null,"abstract":"Capturing interval-valued, as opposed to more conventional point-valued data, offers a potentially efficient method of obtaining richer information in individual responses. In turn, interval-valued data provide a strong foundation for subsequent fuzzy set based modelling-e.g., using the Interval Agreement Approach. In 2019, open-source software (DECSYS) was released to enable digital administration of interval-valued surveys using an ellipse response mode. This study follows on from an appraisal of this software and demonstration of practical value of the approach, reported last year, in one of many potential real-world applications (consumer preference research). A key ambition of ellipse-based interval elicitation is to maximise response efficiency-i.e., minimising workload and complexity in obtaining this richer information. User experience is therefore a vital consideration regarding potential for broader adoption. The present paper documents a direct empirical comparison between interval-valued response elicitation (using ellipses) and a conventional point-valued counterpart (using a Visual Analogue Scale), in terms of user experience during completion of a simple quantitative estimation task. We examine differences in perceived ease-of-use, unnecessary complexity and effective communication of desired responses, as well as overall liking-with positive outcomes for the interval-valued response mode in each case. We also report results of multiple regression analyses examining how the first three variables contribute to participants' overall liking of each response mode, as well as exploring differences driven by potentially important demographic factors (i.e., gender, age & native English speaking).","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130540549","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494549
Marcin Iwanowski, Mateusz Bartosiewicz
Describing the content based on bounding boxes of objects located within the image has recently gained popularity thanks to the fast development of object detection algorithms based on deep learning. Such description, however, does not contain any information on the mutual relations between objects that may be crucial to understand the scene as a whole. In the paper, a method is proposed that extracts, from the set of bounding boxes, a scene description in the form of a list of predicates containing consecutive objects' position, referring them to previously described ones. To estimate bounding boxes' relative position, a fuzzy mutual position matrix is proposed. It contains the complete information on the scene composition stored in fuzzy 2-D position descriptors extracted from fuzzified relative bounding box coordinates by a two-stage fuzzy reasoning process. The descriptors of non-zero membership function values are next considered as potential predicates related to the image content. Their list is ordered using the saliency-based criteria to select the most relevant ones, explaining best the scene composition. From the ordered list, the algorithm extracts the final list of predicates. It contains complete and concise information on the composition of objects within the scene. Some examples of the proposed method illustrate the paper.
{"title":"Describing images using fuzzy mutual position matrix and saliency-based ordering of predicates","authors":"Marcin Iwanowski, Mateusz Bartosiewicz","doi":"10.1109/FUZZ45933.2021.9494549","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494549","url":null,"abstract":"Describing the content based on bounding boxes of objects located within the image has recently gained popularity thanks to the fast development of object detection algorithms based on deep learning. Such description, however, does not contain any information on the mutual relations between objects that may be crucial to understand the scene as a whole. In the paper, a method is proposed that extracts, from the set of bounding boxes, a scene description in the form of a list of predicates containing consecutive objects' position, referring them to previously described ones. To estimate bounding boxes' relative position, a fuzzy mutual position matrix is proposed. It contains the complete information on the scene composition stored in fuzzy 2-D position descriptors extracted from fuzzified relative bounding box coordinates by a two-stage fuzzy reasoning process. The descriptors of non-zero membership function values are next considered as potential predicates related to the image content. Their list is ordered using the saliency-based criteria to select the most relevant ones, explaining best the scene composition. From the ordered list, the algorithm extracts the final list of predicates. It contains complete and concise information on the composition of objects within the scene. Some examples of the proposed method illustrate the paper.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"316 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123232552","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494579
Filipe Santos, J. Sousa, S. Vieira
In this paper, a new approach to the usage of 0-order Autonomous Learning Multi-Model (ALMMo-0) classifiers is proposed. ALMMo-0 classifiers are fully automatic and do not rely on any hyper-parameters. The creation of clouds relies on normalizing data points by their norm, which may remove an important degree of freedom from the data itself. The proposed approach consists of adding the initial radius of the clouds as an hyper-parameter, which makes it possible to skip the normalization step. This approach requires the search for the ideal value of the hyper-parameter. This way, upon training a set of models with different values for the initial radius, the user is expected to be able to choose from several models which range from more accurate to less complex. This approach was tested on three benchmark problems and compared to the results obtained using the original approach. Furthermore, this approach was also tested on a real dataset (Acute Kidney Injury). The obtained results enhance the versatility provided by the proposed method, successfully allowing the user to choose the model that fits better the design demands regarding accuracy, training time, and complexity.
{"title":"A new approach to ALMMo-0 Classifiers: A trade-off between accuracy and complexity","authors":"Filipe Santos, J. Sousa, S. Vieira","doi":"10.1109/FUZZ45933.2021.9494579","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494579","url":null,"abstract":"In this paper, a new approach to the usage of 0-order Autonomous Learning Multi-Model (ALMMo-0) classifiers is proposed. ALMMo-0 classifiers are fully automatic and do not rely on any hyper-parameters. The creation of clouds relies on normalizing data points by their norm, which may remove an important degree of freedom from the data itself. The proposed approach consists of adding the initial radius of the clouds as an hyper-parameter, which makes it possible to skip the normalization step. This approach requires the search for the ideal value of the hyper-parameter. This way, upon training a set of models with different values for the initial radius, the user is expected to be able to choose from several models which range from more accurate to less complex. This approach was tested on three benchmark problems and compared to the results obtained using the original approach. Furthermore, this approach was also tested on a real dataset (Acute Kidney Injury). The obtained results enhance the versatility provided by the proposed method, successfully allowing the user to choose the model that fits better the design demands regarding accuracy, training time, and complexity.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123732858","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494492
J. Fumanal-Idocin, C. Vidaurre, Marisol Gómez, Asier Urio, H. Bustince, M. Papčo, G. Dimuro
Brain-Computer Interfaces based on the analysis of ElectroEncephaloGraphy (EEG) are composed of several elements to process and classify brain input signals. A relevant phase of these systems is the decision making module, in which often the outputs from different classifiers are fused into a single one. In this work, the use of weighted-moderate deviation based functions is proposed to improve the Enhanced-Multimodal Fusion BCI Framework (EMF) decision making phase. Moderate Deviation-based aggregation functions (MDs) allow us to choose the best value to aggregate a vector of points involving a moderate deviation function. Using a weighted MD, the relative importance of each dimension in the multi-dimensional aggregated data set can also be taken into account. By applying these functions in the EMF, each one of the different brain signals can be weighted according to their importance. Moreover, using automatic differentiation, it is possible to optimize them for the present problem.
{"title":"Optimizing a Weighted Moderate Deviation for Motor Imagery Brain Computer Interfaces","authors":"J. Fumanal-Idocin, C. Vidaurre, Marisol Gómez, Asier Urio, H. Bustince, M. Papčo, G. Dimuro","doi":"10.1109/FUZZ45933.2021.9494492","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494492","url":null,"abstract":"Brain-Computer Interfaces based on the analysis of ElectroEncephaloGraphy (EEG) are composed of several elements to process and classify brain input signals. A relevant phase of these systems is the decision making module, in which often the outputs from different classifiers are fused into a single one. In this work, the use of weighted-moderate deviation based functions is proposed to improve the Enhanced-Multimodal Fusion BCI Framework (EMF) decision making phase. Moderate Deviation-based aggregation functions (MDs) allow us to choose the best value to aggregate a vector of points involving a moderate deviation function. Using a weighted MD, the relative importance of each dimension in the multi-dimensional aggregated data set can also be taken into account. By applying these functions in the EMF, each one of the different brain signals can be weighted according to their importance. Moreover, using automatic differentiation, it is possible to optimize them for the present problem.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124905071","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494494
R. Guillaume, A. Kasperski, P. Zieliński
In this paper a robust optimization problem with uncertain objective function is considered. The uncertainty is modeled by specifying a scenario set, containing a finite number of objective function coefficients, called scenarios. Additional knowledge in scenario set can be represented by using a mass function defined on the power set of scenarios. This mass function defines a belief function, which in turn induces a family of probability distributions in scenario set. One can then use a generalized Hurwicz criterion, i.e. a convex combination of the upper and lower expectations, to solve the uncertain problem. Recently, possibility theory has been applied to extend the model of uncertainty based on belief functions. Namely, belief function can be induced by a random fuzzy set. In this paper we show how this generalized model can be applied to robust optimization.
{"title":"Robust optimization with scenarios using random fuzzy sets","authors":"R. Guillaume, A. Kasperski, P. Zieliński","doi":"10.1109/FUZZ45933.2021.9494494","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494494","url":null,"abstract":"In this paper a robust optimization problem with uncertain objective function is considered. The uncertainty is modeled by specifying a scenario set, containing a finite number of objective function coefficients, called scenarios. Additional knowledge in scenario set can be represented by using a mass function defined on the power set of scenarios. This mass function defines a belief function, which in turn induces a family of probability distributions in scenario set. One can then use a generalized Hurwicz criterion, i.e. a convex combination of the upper and lower expectations, to solve the uncertain problem. Recently, possibility theory has been applied to extend the model of uncertainty based on belief functions. Namely, belief function can be induced by a random fuzzy set. In this paper we show how this generalized model can be applied to robust optimization.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116778543","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494471
Dorukhan Erdem, T. Kumbasar
Fuzzy Logic Systems (FLSs), especially Interval Type-2 (IT2) ones, are proven to achieve good results in various tasks, including classification problems. However, IT2-FLSs suffer from the curse of dimensionality problem, just like its Type-1 (T1) counterparts, and also training complexity since IT2-FLS have a large number of learnable parameters when compared to T1-FLSs. Deep learning (DL) architectures on the other hand can handle large learnable parameter sets for good generalizability but have their disadvantages. In this study, we present DL based approach with knowledge distillation for IT2-FLSs which transfers the generalizability features of deep models into IT2-FLS and increases its learning performance significantly by eliminating the problems that may arise from large input sizes and high rule counts. We present in detail the proposed approach with parameterization tricks so that the training of IT2-FLS can be accomplished straightforwardly within the widely employed DL frameworks without violating the definitions of IT2-FSs. We present comparative analysis to show the benefits of the inclusion knowledge distillation in the learning of IT2-FLSs with respect to rule number and input dimension size.
{"title":"Enhancing the Learning of Interval Type-2 Fuzzy Classifiers with Knowledge Distillation","authors":"Dorukhan Erdem, T. Kumbasar","doi":"10.1109/FUZZ45933.2021.9494471","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494471","url":null,"abstract":"Fuzzy Logic Systems (FLSs), especially Interval Type-2 (IT2) ones, are proven to achieve good results in various tasks, including classification problems. However, IT2-FLSs suffer from the curse of dimensionality problem, just like its Type-1 (T1) counterparts, and also training complexity since IT2-FLS have a large number of learnable parameters when compared to T1-FLSs. Deep learning (DL) architectures on the other hand can handle large learnable parameter sets for good generalizability but have their disadvantages. In this study, we present DL based approach with knowledge distillation for IT2-FLSs which transfers the generalizability features of deep models into IT2-FLS and increases its learning performance significantly by eliminating the problems that may arise from large input sizes and high rule counts. We present in detail the proposed approach with parameterization tricks so that the training of IT2-FLS can be accomplished straightforwardly within the widely employed DL frameworks without violating the definitions of IT2-FSs. We present comparative analysis to show the benefits of the inclusion knowledge distillation in the learning of IT2-FLSs with respect to rule number and input dimension size.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116839140","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494540
Shreyas J. Upasane, H. Hagras, M. Anisi, Stuart Savill, Ian J. Taylor, Kostas Manousakis
The role of maintenance in modern manufacturing systems is becoming a more significant contributor to organizational benefit. World-class enterprises are pushing forward with “predict-and prevent” maintenance instead of embracing the drawbacks of reactive maintenance (or a “fail-and fix” approach). The advancement towards Artificial Intelligence (AI), Internet of Things (IoT) and cloud computing has led to a shift in maintenance paradigms with the rising interest in Machine Learning (ML) and in particular deep learning. However, opaque box AI models are complex and difficult to understand and explain to the lay user. This limits the use of these models in predictive maintenance where it is crucial to understand and analyze the model before deployment and it is imperative to understand the logic behind any given decision. This paper introduces a Type-2 Fuzzy Logic System (FLS) optimized by the Big-Bang Big-Crunch algorithm that allows maximizing the interpretability of a model as well as its prediction accuracy for the faults which may occur in future. We tested the proposed type-2 FLS model on water pumps where data was collected in real-time by our proprietary hardware deployed at Aquatronic Group Management Plc. The observations indicate that the proposed system provides a highly interpretable and accurate model for predicting the faults in equipment for building services, process and water industries. The system predictions are used to understand why a particular fault may occur, leading to improved and better-informed service visits for the customers thus reducing the disruptions faced due to equipment failures.
维护在现代制造系统中的作用正在成为组织效益的重要贡献者。世界级的企业正在推进“预测和预防”维护，而不是接受被动维护的缺点(或“故障和修复”方法)。人工智能(AI)、物联网(IoT)和云计算的发展导致了维护范式的转变，人们对机器学习(ML)，特别是深度学习的兴趣日益浓厚。然而，不透明的盒子人工智能模型是复杂的，很难理解和解释给外行用户。这限制了这些模型在预测性维护中的使用，在预测性维护中，在部署之前理解和分析模型是至关重要的，并且必须理解任何给定决策背后的逻辑。本文介绍了一种采用大爆炸大压缩算法优化的2型模糊逻辑系统(FLS)，该系统可以最大限度地提高模型的可解释性和对未来可能发生的故障的预测精度。我们在水泵上测试了type-2 FLS模型，通过Aquatronic Group Management Plc部署的专有硬件实时收集数据。观察结果表明，所提出的系统为建筑服务、过程和水工业设备故障预测提供了一个高度可解释和准确的模型。系统预测用于了解特定故障可能发生的原因，从而为客户提供改进和更明智的服务访问，从而减少因设备故障而面临的中断。
{"title":"A Big Bang-Big Crunch Type-2 Fuzzy Logic System for Explainable Predictive Maintenance","authors":"Shreyas J. Upasane, H. Hagras, M. Anisi, Stuart Savill, Ian J. Taylor, Kostas Manousakis","doi":"10.1109/FUZZ45933.2021.9494540","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494540","url":null,"abstract":"The role of maintenance in modern manufacturing systems is becoming a more significant contributor to organizational benefit. World-class enterprises are pushing forward with “predict-and prevent” maintenance instead of embracing the drawbacks of reactive maintenance (or a “fail-and fix” approach). The advancement towards Artificial Intelligence (AI), Internet of Things (IoT) and cloud computing has led to a shift in maintenance paradigms with the rising interest in Machine Learning (ML) and in particular deep learning. However, opaque box AI models are complex and difficult to understand and explain to the lay user. This limits the use of these models in predictive maintenance where it is crucial to understand and analyze the model before deployment and it is imperative to understand the logic behind any given decision. This paper introduces a Type-2 Fuzzy Logic System (FLS) optimized by the Big-Bang Big-Crunch algorithm that allows maximizing the interpretability of a model as well as its prediction accuracy for the faults which may occur in future. We tested the proposed type-2 FLS model on water pumps where data was collected in real-time by our proprietary hardware deployed at Aquatronic Group Management Plc. The observations indicate that the proposed system provides a highly interpretable and accurate model for predicting the faults in equipment for building services, process and water industries. The system predictions are used to understand why a particular fault may occur, leading to improved and better-informed service visits for the customers thus reducing the disruptions faced due to equipment failures.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114448037","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494489
V. Noghin, O. Baskov
A multicriteria choice problem is considered. The setting of this problem includes three objects, namely, a set of feasible alternatives, a numerical vector criterion, and a decision maker's binary strict preference relation. The Edgeworth — Pareto principle is a fundamental instrument to solve multi-criteria problems. Previously, the validity of this principle was established in the case of a crisp as well as a type-1 fuzzy preference relation. We assume that the preference relation is a type-2 fuzzy relation. Under two reasonable axioms the Edgeworth—Pareto principle is established. In accordance with the first axiom, an alternative not chosen in a pair should not be selected from the whole set of feasible alternatives. The second axiom is the Pareto axiom, which provides greater preference for those alternatives that have larger (smaller) values of one or more criteria.
{"title":"On Multicriteria Choice Based on Type-2 Fuzzy Preference Relation: an Axiomatic Approach","authors":"V. Noghin, O. Baskov","doi":"10.1109/FUZZ45933.2021.9494489","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494489","url":null,"abstract":"A multicriteria choice problem is considered. The setting of this problem includes three objects, namely, a set of feasible alternatives, a numerical vector criterion, and a decision maker's binary strict preference relation. The Edgeworth — Pareto principle is a fundamental instrument to solve multi-criteria problems. Previously, the validity of this principle was established in the case of a crisp as well as a type-1 fuzzy preference relation. We assume that the preference relation is a type-2 fuzzy relation. Under two reasonable axioms the Edgeworth—Pareto principle is established. In accordance with the first axiom, an alternative not chosen in a pair should not be selected from the whole set of feasible alternatives. The second axiom is the Pareto axiom, which provides greater preference for those alternatives that have larger (smaller) values of one or more criteria.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130265861","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 : 2021-07-11DOI: 10.1109/FUZZ45933.2021.9494479
Guoliang Feng, Wei Lu, Jianhua Yang
The change of amplitude and frequency result in a variety of variation modality of time series in the universe. It is difficult to describe the variation features of time series exactly relying solely on a single simulating model. To overcome this limitation, a new prediction model using fuzzy cognitive maps is proposed based on partitioning strategies. Initially, fuzzy c-mean clustering is adopted to partition time series into several sub-sequences. Consequently, each partition has its corresponding sequences. Subsequently these sub-sequences are used to constructed fuzzy cognitive maps models respectively. Finally, the fuzzy cognitive maps models are merged by fuzzy rules. The constructed model is not only performing well in numerical prediction but also has interpretability. The experimental results show that the model based on partition strategy is superior to the single.
{"title":"Time Series Modeling with Fuzzy Cognitive Maps based on Partitioning Strategies","authors":"Guoliang Feng, Wei Lu, Jianhua Yang","doi":"10.1109/FUZZ45933.2021.9494479","DOIUrl":"https://doi.org/10.1109/FUZZ45933.2021.9494479","url":null,"abstract":"The change of amplitude and frequency result in a variety of variation modality of time series in the universe. It is difficult to describe the variation features of time series exactly relying solely on a single simulating model. To overcome this limitation, a new prediction model using fuzzy cognitive maps is proposed based on partitioning strategies. Initially, fuzzy c-mean clustering is adopted to partition time series into several sub-sequences. Consequently, each partition has its corresponding sequences. Subsequently these sub-sequences are used to constructed fuzzy cognitive maps models respectively. Finally, the fuzzy cognitive maps models are merged by fuzzy rules. The constructed model is not only performing well in numerical prediction but also has interpretability. The experimental results show that the model based on partition strategy is superior to the single.","PeriodicalId":151289,"journal":{"name":"2021 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130093777","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：