Pub Date : 2021-11-01DOI: 10.23919/fusion49465.2021.9627035
Charlotte Jacobé de Naurois, C. Laudy
We previously developped InSyTo, a framework for soft and semantic information fusion and management relying on the conceptual graph formalism and ontologies. This framework was used in many projects. However, if the framework was well received by end-users, they highlighted an urgent need for traceability within the soft information fusion process. In this paper we propose an approach to provide traceability built-in capabilities to InSyTo. The approach relies on the use of conceptual graphs in order to express the history of each elementary piece of information as a lineage graph. The lineage graph contains all the historical information concerning the initial sources of each elementary information item, as well as the fusion operations that were applied on them. The main advantage of this new development is the ability of having a trustworthy framework and thus let the end-users know what, why and how somethings happens during the information process.
{"title":"Handling Traceability in Graph Fusion for a Trustworthy Framework","authors":"Charlotte Jacobé de Naurois, C. Laudy","doi":"10.23919/fusion49465.2021.9627035","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9627035","url":null,"abstract":"We previously developped InSyTo, a framework for soft and semantic information fusion and management relying on the conceptual graph formalism and ontologies. This framework was used in many projects. However, if the framework was well received by end-users, they highlighted an urgent need for traceability within the soft information fusion process. In this paper we propose an approach to provide traceability built-in capabilities to InSyTo. The approach relies on the use of conceptual graphs in order to express the history of each elementary piece of information as a lineage graph. The lineage graph contains all the historical information concerning the initial sources of each elementary information item, as well as the fusion operations that were applied on them. The main advantage of this new development is the ability of having a trustworthy framework and thus let the end-users know what, why and how somethings happens during the information process.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"367 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133933014","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9627061
Avi Chawla, Nidhi Mulay, Vikas Bishnoi, Gaurav Dhama
Large Transformer based models have provided state-of-the-art results on a variety of Natural Language Processing (NLP) tasks. While these Transformer models perform exceptionally well on a wide range of NLP tasks, their usage in Sequence Labeling has been mostly muted. Although pretrained Transformer models such as BERT and XLNet have been successfully employed as input representation, the use of the Transformer model as a context encoder for sequence labeling is still minimal, and most recent works still use recurrent architecture as the context encoder. In this paper, we compare the performance of the Transformer and Recurrent architecture as context encoders on the Named Entity Recognition (NER) task. We vary the character-level representation module from the previously proposed NER models in literature and show how the modification can improve the NER model’s performance. We also explore data augmentation as a method for enhancing their performance. Experimental results on three NER datasets show that our proposed techniques established a new state-of-the-art using the Transformer Encoder over the previously proposed models in the literature using only non-contextualized embeddings.
{"title":"Improving the performance of Transformer Context Encoders for NER","authors":"Avi Chawla, Nidhi Mulay, Vikas Bishnoi, Gaurav Dhama","doi":"10.23919/fusion49465.2021.9627061","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9627061","url":null,"abstract":"Large Transformer based models have provided state-of-the-art results on a variety of Natural Language Processing (NLP) tasks. While these Transformer models perform exceptionally well on a wide range of NLP tasks, their usage in Sequence Labeling has been mostly muted. Although pretrained Transformer models such as BERT and XLNet have been successfully employed as input representation, the use of the Transformer model as a context encoder for sequence labeling is still minimal, and most recent works still use recurrent architecture as the context encoder. In this paper, we compare the performance of the Transformer and Recurrent architecture as context encoders on the Named Entity Recognition (NER) task. We vary the character-level representation module from the previously proposed NER models in literature and show how the modification can improve the NER model’s performance. We also explore data augmentation as a method for enhancing their performance. Experimental results on three NER datasets show that our proposed techniques established a new state-of-the-art using the Transformer Encoder over the previously proposed models in the literature using only non-contextualized embeddings.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"397 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116674294","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626831
Hongpo Fu, Yong-mei Cheng, Cheng Cheng
In the study of the state estimation for the systems with unknown time-varying non-Gaussian noises, the existing robust Kalman filters (RKFs) perform well. However, the calculation loads of these RKFs usually are large and their performance is easily affected by the roughly preselected initial process noise covariance matrix (PNCM). To solve the problems, a new RKF is proposed. Firstly, a Gaussian-inverse Gamma mixture distribution is developed to model the inaccurate noises and a simple hierarchical Gaussian (HG) model is constructed. Then, the expectation-maximization (EM) method is applied to realize the adaptive adjustment of the prior scale matrix of the prediction error covariance. Based on the HG model and EM, a robust KF is derived, where the variational Bayesian (VB) approach is used to jointly estimate model parameters and an alternate iteration method is employed to reduce the computation time. Finally, our filter performance is tested. Compared with the existing state-of-the-art robust filters, the proposed filter has slightly better estimation accuracy and significantly less computation load. Meanwhile, the filter performance is almost not affected by the selection accuracy of initial PNCM.
{"title":"A Gaussian-inverse Gamma mixture Distributions and Expectation-Maximization Based Robust Kalman Filter","authors":"Hongpo Fu, Yong-mei Cheng, Cheng Cheng","doi":"10.23919/fusion49465.2021.9626831","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626831","url":null,"abstract":"In the study of the state estimation for the systems with unknown time-varying non-Gaussian noises, the existing robust Kalman filters (RKFs) perform well. However, the calculation loads of these RKFs usually are large and their performance is easily affected by the roughly preselected initial process noise covariance matrix (PNCM). To solve the problems, a new RKF is proposed. Firstly, a Gaussian-inverse Gamma mixture distribution is developed to model the inaccurate noises and a simple hierarchical Gaussian (HG) model is constructed. Then, the expectation-maximization (EM) method is applied to realize the adaptive adjustment of the prior scale matrix of the prediction error covariance. Based on the HG model and EM, a robust KF is derived, where the variational Bayesian (VB) approach is used to jointly estimate model parameters and an alternate iteration method is employed to reduce the computation time. Finally, our filter performance is tested. Compared with the existing state-of-the-art robust filters, the proposed filter has slightly better estimation accuracy and significantly less computation load. Meanwhile, the filter performance is almost not affected by the selection accuracy of initial PNCM.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123503033","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626988
C. Musso, F. Dambreville, C. Chahbazian
Passive target estimation is a widely investigated problem of practical interest for which particle filters represent a popular class of methods. We propose an adaptation of the Laplace Particle Filter applied to angle-only navigation using landmarks. In this specific context, a high number of aiding landmarks or features could be hard to handle in terms of computational cost. Hence, this paper introduces a Cross-entropy algorithm that selects landmarks having a high contribution to the state estimation. This parsimonious approach reduces the resources required for navigation systems while holding a good accuracy. These methods are discussed through numerical results on an Angle-only navigation scenario.
{"title":"Filtering and sensor optimization applied to angle-only navigation","authors":"C. Musso, F. Dambreville, C. Chahbazian","doi":"10.23919/fusion49465.2021.9626988","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626988","url":null,"abstract":"Passive target estimation is a widely investigated problem of practical interest for which particle filters represent a popular class of methods. We propose an adaptation of the Laplace Particle Filter applied to angle-only navigation using landmarks. In this specific context, a high number of aiding landmarks or features could be hard to handle in terms of computational cost. Hence, this paper introduces a Cross-entropy algorithm that selects landmarks having a high contribution to the state estimation. This parsimonious approach reduces the resources required for navigation systems while holding a good accuracy. These methods are discussed through numerical results on an Angle-only navigation scenario.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125242609","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626922
M. J. Ransom, M. Hernandez, J. Ralph, S. Maskell
This paper is concerned with the implementation of track-before-detect (TkBD) algorithms for a range of single-target multi-sensor scenarios with only intermittently visible targets. Visible targets generate measurements from sensors characterised by data rate and clutter density. Bernoulli filters implementing multiple hypothesis tracking (MHT) strategies are deployed to infer both the target location and existence probability. Various Bernoulli filter configurations are compared, including integrated probabilistic data association filters (IPDAF) and integrated expected likelihood particle filters (IELPF) using both prior and Gaussian mixture proposal distributions for the latter. Performance is evaluated against the clutter density in scenarios featuring one low data rate active sensor or two sensors, complimenting the former with a high data rate passive sensor with opposing measurement resolutions. The performance measures used are the area under the receiver operating characteristic (ROC) curve, localisation root mean squared error (RMSE) compared with the posterior Cramér-Rao lower bound (PCRLB), and computation time. Simulation results show that Kalman filters provide an effective solution at low computational expense in less cluttered and comparatively easy scenarios, whereas particle filters implementing Gaussian mixture proposal distributions provide performance benefits relative to computational costs as scenarios become more cluttered and comparatively challenging.
{"title":"Track-before-detect Bernoulli filters for combining passive and active sensors","authors":"M. J. Ransom, M. Hernandez, J. Ralph, S. Maskell","doi":"10.23919/fusion49465.2021.9626922","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626922","url":null,"abstract":"This paper is concerned with the implementation of track-before-detect (TkBD) algorithms for a range of single-target multi-sensor scenarios with only intermittently visible targets. Visible targets generate measurements from sensors characterised by data rate and clutter density. Bernoulli filters implementing multiple hypothesis tracking (MHT) strategies are deployed to infer both the target location and existence probability. Various Bernoulli filter configurations are compared, including integrated probabilistic data association filters (IPDAF) and integrated expected likelihood particle filters (IELPF) using both prior and Gaussian mixture proposal distributions for the latter. Performance is evaluated against the clutter density in scenarios featuring one low data rate active sensor or two sensors, complimenting the former with a high data rate passive sensor with opposing measurement resolutions. The performance measures used are the area under the receiver operating characteristic (ROC) curve, localisation root mean squared error (RMSE) compared with the posterior Cramér-Rao lower bound (PCRLB), and computation time. Simulation results show that Kalman filters provide an effective solution at low computational expense in less cluttered and comparatively easy scenarios, whereas particle filters implementing Gaussian mixture proposal distributions provide performance benefits relative to computational costs as scenarios become more cluttered and comparatively challenging.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126180931","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626911
Jonas Åsnes Sagild, Audun Gullikstad Hem, E. Brekke
Standard hypothesis tests for track-to-track association depend on the state estimates and covariances of the individual tracks. Unfortunately, covariances are not always available from the individual tracking systems. An alternative approach that can be used in such cases is a counting technique, where the number of good matches is used as a test statistic. In this paper, we compare the counting technique with a conventional hypothesis test in simulations for a fusion system designed to fuse maritime radar tracks with tracks from the automatic identification system. Since the data association of the radar tracking system inevitably makes it nontrivial to decide on a ground truth, we also propose a ground truth assessment scheme using a sliding window approach. The results indicate that the counting technique performs at par with the hypothesis test under certain tracking conditions. If an initialization time of several seconds is allowed, the counting technique may under certain conditions outperform the hypothesis test in terms of true-positive rate and false-positive rate.
{"title":"Counting Technique versus Single-Time Test for Track-to-Track Association","authors":"Jonas Åsnes Sagild, Audun Gullikstad Hem, E. Brekke","doi":"10.23919/fusion49465.2021.9626911","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626911","url":null,"abstract":"Standard hypothesis tests for track-to-track association depend on the state estimates and covariances of the individual tracks. Unfortunately, covariances are not always available from the individual tracking systems. An alternative approach that can be used in such cases is a counting technique, where the number of good matches is used as a test statistic. In this paper, we compare the counting technique with a conventional hypothesis test in simulations for a fusion system designed to fuse maritime radar tracks with tracks from the automatic identification system. Since the data association of the radar tracking system inevitably makes it nontrivial to decide on a ground truth, we also propose a ground truth assessment scheme using a sliding window approach. The results indicate that the counting technique performs at par with the hypothesis test under certain tracking conditions. If an initialization time of several seconds is allowed, the counting technique may under certain conditions outperform the hypothesis test in terms of true-positive rate and false-positive rate.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124604333","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626913
Peng Liu, Z. Duan
3D multi-object tracking (MOT) is a crucial part in the field of autonomous driving. Thanks to the recent advances in deep-learning-based detector, tracking-by-detection paradigm has become popular in 3D MOT, which consists of a front-end object detector and a back-end tracker. However, most existing methods only focus on the performance on particular data sets, ignoring the adaptiveness of the tracking algorithm to dynamically changing driving environment. Based on this, we design an adaptive 3D MOT algorithm, which can adapt its behavior to the complex changing environments in real driving scenarios. The system first utilizes a pre-trained 3D detector to produce the observations (detections) for the current frame. Then, a state estimator based on interacting multiple model (IMM), which takes the statistics of the data set into account and switches its state dynamically, provides the adaptive state estimation for target tracking. Experiments show that our algorithm can improve the performance of single-model-based methods, and adapt its behavior dynamically on nuScenes data set.
{"title":"An IMM-Enabled Adaptive 3D Multi-Object Tracker for Autonomous Driving","authors":"Peng Liu, Z. Duan","doi":"10.23919/fusion49465.2021.9626913","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626913","url":null,"abstract":"3D multi-object tracking (MOT) is a crucial part in the field of autonomous driving. Thanks to the recent advances in deep-learning-based detector, tracking-by-detection paradigm has become popular in 3D MOT, which consists of a front-end object detector and a back-end tracker. However, most existing methods only focus on the performance on particular data sets, ignoring the adaptiveness of the tracking algorithm to dynamically changing driving environment. Based on this, we design an adaptive 3D MOT algorithm, which can adapt its behavior to the complex changing environments in real driving scenarios. The system first utilizes a pre-trained 3D detector to produce the observations (detections) for the current frame. Then, a state estimator based on interacting multiple model (IMM), which takes the statistics of the data set into account and switches its state dynamically, provides the adaptive state estimation for target tracking. Experiments show that our algorithm can improve the performance of single-model-based methods, and adapt its behavior dynamically on nuScenes data set.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121534424","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9627017
B. K. Almentero, Jiye Li, C. Besse
Inventory represents the largest asset in pharmacy products distribution. Forecasting pharmacy purchases is essential to keep an effective stock balancing supply and demand besides minimizing costs. In this work, we investigate how to forecast product purchases for a pharmaceutical distributor. The data contains inventory sale histories for more than 10 thousand active products during the past 15 years. We discuss challenges on data preprocessing of the pharmacy data including cleaning, feature constructions and selections, as well as processing data during the COVID period. We experimented on different machine learning and deep learning neural network models to predict future purchases for each product, including classical Seasonal Autoregressive Integrated Moving Average (SARIMA), Prophet from Facebook, linear regression, Random Forest, XGBoost and Long Short-Term Memory (LSTM). We demonstrate that a carefully designed SARIMA model outperformed the others on the task, and weekly prediction models perform better than daily predictions.
{"title":"Forecasting pharmacy purchases orders","authors":"B. K. Almentero, Jiye Li, C. Besse","doi":"10.23919/fusion49465.2021.9627017","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9627017","url":null,"abstract":"Inventory represents the largest asset in pharmacy products distribution. Forecasting pharmacy purchases is essential to keep an effective stock balancing supply and demand besides minimizing costs. In this work, we investigate how to forecast product purchases for a pharmaceutical distributor. The data contains inventory sale histories for more than 10 thousand active products during the past 15 years. We discuss challenges on data preprocessing of the pharmacy data including cleaning, feature constructions and selections, as well as processing data during the COVID period. We experimented on different machine learning and deep learning neural network models to predict future purchases for each product, including classical Seasonal Autoregressive Integrated Moving Average (SARIMA), Prophet from Facebook, linear regression, Random Forest, XGBoost and Long Short-Term Memory (LSTM). We demonstrate that a carefully designed SARIMA model outperformed the others on the task, and weekly prediction models perform better than daily predictions.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122663002","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626981
Makhetha Molefi, E. Markus, A. Abu-Mahfouz
A Conformal Strongly Coupled Magnetic Resonance (CSCMR) wireless power transfer (WPT) system is a small footprint technology suitable for applications such as small low power sensors and implantable medical devices. These applications require specific WPT systems with certain physical dimensions that complement the size of the device. The design of these systems can be complex and require intense computational resources and long simulation times to conceptualise the optimal WPT system. This paper discusses the system architecture for CSCMR-WPT model. A quicker mathematical analysis to estimate the optimal CSCMR-WPT resonator loops and source/load loops is shown. The results confirm that this method can lead to quicker conceptualisation of a WPT model.
{"title":"Accelerated Design of a Conformal Strongly Coupled Magnetic Resonance Wireless Power Transfer","authors":"Makhetha Molefi, E. Markus, A. Abu-Mahfouz","doi":"10.23919/fusion49465.2021.9626981","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626981","url":null,"abstract":"A Conformal Strongly Coupled Magnetic Resonance (CSCMR) wireless power transfer (WPT) system is a small footprint technology suitable for applications such as small low power sensors and implantable medical devices. These applications require specific WPT systems with certain physical dimensions that complement the size of the device. The design of these systems can be complex and require intense computational resources and long simulation times to conceptualise the optimal WPT system. This paper discusses the system architecture for CSCMR-WPT model. A quicker mathematical analysis to estimate the optimal CSCMR-WPT resonator loops and source/load loops is shown. The results confirm that this method can lead to quicker conceptualisation of a WPT model.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122760521","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 : 2021-11-01DOI: 10.23919/fusion49465.2021.9626835
J. Dezert, A. Bouchard, M. Buguet
The objective of this paper is to present a general methodology for storm risk assessment and prediction based on several physical criteria thanks to the belief functions framework to deal with conflicting meteorological information. For this, we adapt the Soft ELECTRE TRI (SET) approach to this storm context and we show how to use it on outputs of atmospheric forecast model, given an estimate of the state of the atmosphere in a future time. This work could also serve as a benchmark for other methods dealing with multi-criteria decision-making (MCDM) support and conflicting information fusion.
本文的目的是提出一种基于几种物理标准的风暴风险评估和预测的一般方法，这要感谢信念函数框架来处理相互冲突的气象信息。为此，我们将Soft ELECTRE TRI (SET)方法应用于该风暴背景，并展示了如何在给定未来大气状态估计的情况下将其用于大气预测模型的输出。该工作也可以为其他处理多准则决策支持和冲突信息融合的方法提供参考。
{"title":"Multi-Criteria Information Fusion for Storm Prediction Based on Belief Functions","authors":"J. Dezert, A. Bouchard, M. Buguet","doi":"10.23919/fusion49465.2021.9626835","DOIUrl":"https://doi.org/10.23919/fusion49465.2021.9626835","url":null,"abstract":"The objective of this paper is to present a general methodology for storm risk assessment and prediction based on several physical criteria thanks to the belief functions framework to deal with conflicting meteorological information. For this, we adapt the Soft ELECTRE TRI (SET) approach to this storm context and we show how to use it on outputs of atmospheric forecast model, given an estimate of the state of the atmosphere in a future time. This work could also serve as a benchmark for other methods dealing with multi-criteria decision-making (MCDM) support and conflicting information fusion.","PeriodicalId":226850,"journal":{"name":"2021 IEEE 24th International Conference on Information Fusion (FUSION)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122989899","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：