Pub Date : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00270
Marco Dossena, Christopher Irwin, L. Portinale
Graph based recommendation strategies are recently gaining momentum in connection with the availability of new Graph Neural Network (GNN) architectures. In fact, the interactions between users and products in a recommender system can be naturally expressed in terms of a bipartite graph, where nodes corresponding to users are connected with nodes corresponding to products trough edges representing a user action on the product (usually a purchase). GNNs can then be exploited and trained in order to predict the existence of a specific edge between unconnected users and products, highlighting the interest for a potential purchase of a given product by a given user. In the present paper, we will present an experimental analysis of different GNN architectures in the context of recommender systems. We analyze the impact of different kind of layers such as convolutional, attentional and message-passing, as well as the influence of different embedding size on the performance on the link prediction task. We will also examine the behavior of two of such architectures (the ones relying on the presence of node features) with respect to both transductive and inductive situations.
{"title":"Graph-based Recommendation using Graph Neural Networks","authors":"Marco Dossena, Christopher Irwin, L. Portinale","doi":"10.1109/ICMLA55696.2022.00270","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00270","url":null,"abstract":"Graph based recommendation strategies are recently gaining momentum in connection with the availability of new Graph Neural Network (GNN) architectures. In fact, the interactions between users and products in a recommender system can be naturally expressed in terms of a bipartite graph, where nodes corresponding to users are connected with nodes corresponding to products trough edges representing a user action on the product (usually a purchase). GNNs can then be exploited and trained in order to predict the existence of a specific edge between unconnected users and products, highlighting the interest for a potential purchase of a given product by a given user. In the present paper, we will present an experimental analysis of different GNN architectures in the context of recommender systems. We analyze the impact of different kind of layers such as convolutional, attentional and message-passing, as well as the influence of different embedding size on the performance on the link prediction task. We will also examine the behavior of two of such architectures (the ones relying on the presence of node features) with respect to both transductive and inductive situations.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121500993","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00265
Jessica An
Alzheimer's disease is a progressive neurologic disorder that affects millions of elderly people worldwide. Most affected patients are not formally diagnosed due to the complexity of the disease and the lack of definitive diagnostic tools. Machine learning algorithms are powerful in deciphering complex data patterns. This study applied and evaluated a comprehensive set of nine machine learning techniques in detecting Alzheimer's disease. The model training and testing utilized clinical and brain magnetic resonance imaging features from The Open Access Series of Imaging Studies (OASIS) of Alzheimer's disease. The input data include ordinal data such as cognitive scores and numerical data of imaging measurements. To predict Alzheimer's disease, multiple types of supervised machine learning algorithms were trained, including CatBoost, logistic regression, decision tree, random forest, Naive Bayes, SVM, gradient boosting, XGBoost, and AdaBoost. A set of model performance metrics demonstrated that most algorithms were able to perform very well with high accuracy (92-96% in a longitudinal dataset). The models using CatBoost, SVM and decision tree performed the best. The results of this study suggest that ML algorithms combining clinical cognitive assessment and brain MRI images can assist and improve Alzheimer's disease diagnosis.
{"title":"Using CatBoost and Other Supervised Machine Learning Algorithms to Predict Alzheimer's Disease","authors":"Jessica An","doi":"10.1109/ICMLA55696.2022.00265","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00265","url":null,"abstract":"Alzheimer's disease is a progressive neurologic disorder that affects millions of elderly people worldwide. Most affected patients are not formally diagnosed due to the complexity of the disease and the lack of definitive diagnostic tools. Machine learning algorithms are powerful in deciphering complex data patterns. This study applied and evaluated a comprehensive set of nine machine learning techniques in detecting Alzheimer's disease. The model training and testing utilized clinical and brain magnetic resonance imaging features from The Open Access Series of Imaging Studies (OASIS) of Alzheimer's disease. The input data include ordinal data such as cognitive scores and numerical data of imaging measurements. To predict Alzheimer's disease, multiple types of supervised machine learning algorithms were trained, including CatBoost, logistic regression, decision tree, random forest, Naive Bayes, SVM, gradient boosting, XGBoost, and AdaBoost. A set of model performance metrics demonstrated that most algorithms were able to perform very well with high accuracy (92-96% in a longitudinal dataset). The models using CatBoost, SVM and decision tree performed the best. The results of this study suggest that ML algorithms combining clinical cognitive assessment and brain MRI images can assist and improve Alzheimer's disease diagnosis.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126480998","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00034
Feiyang Cai, Zhenkai Zhang, Jie Liu
Deep neural networks have demonstrated prominent capacities for image classification tasks in a closed set setting, where the test data come from the same distribution as the training data. However, in a more realistic open set scenario, traditional classifiers with incomplete knowledge cannot tackle test data that are not from the training classes. Open set recognition (OSR) aims to address this problem by both identifying unknown classes and distinguishing known classes simultaneously. In this paper, we propose a novel approach to OSR that is based on the vision transformer (ViT) technique. Specifically, our approach employs two separate training stages. First, a ViT model is trained to perform closed set classification. Then, an additional detection head is attached to the embedded features extracted by the ViT, trained to force the representations of known data to class-specific clusters compactly. Test examples are identified as known or unknown based on their distance to the cluster centers. To the best of our knowledge, this is the first time to leverage ViT for the purpose of OSR, and our extensive evaluation against several OSR benchmark datasets reveals that our approach significantly outperforms other baseline methods and obtains new state-of-the-art performance.
{"title":"A Vision Transformer Architecture for Open Set Recognition","authors":"Feiyang Cai, Zhenkai Zhang, Jie Liu","doi":"10.1109/ICMLA55696.2022.00034","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00034","url":null,"abstract":"Deep neural networks have demonstrated prominent capacities for image classification tasks in a closed set setting, where the test data come from the same distribution as the training data. However, in a more realistic open set scenario, traditional classifiers with incomplete knowledge cannot tackle test data that are not from the training classes. Open set recognition (OSR) aims to address this problem by both identifying unknown classes and distinguishing known classes simultaneously. In this paper, we propose a novel approach to OSR that is based on the vision transformer (ViT) technique. Specifically, our approach employs two separate training stages. First, a ViT model is trained to perform closed set classification. Then, an additional detection head is attached to the embedded features extracted by the ViT, trained to force the representations of known data to class-specific clusters compactly. Test examples are identified as known or unknown based on their distance to the cluster centers. To the best of our knowledge, this is the first time to leverage ViT for the purpose of OSR, and our extensive evaluation against several OSR benchmark datasets reveals that our approach significantly outperforms other baseline methods and obtains new state-of-the-art performance.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121790448","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00138
A. Alizadeh, Pooya Tavallali, Vahid Behzadan, A. Ranganath, Mukesh Singhal
Synthetic Reduced Nearest Neighbor is a nearest neighbor model that is constrained on synthetic samples (i.e., prototypes). The body of work on such models includes proposals for improving the interpretability and optimization of SRNN models using expectation maximization. Motivated by the promise of this paradigm, we propose a novel Expectation Maximization approach for Synthetic Reduced Nearest Neighbors leveraging neural networks. Furthermore, we compare the performance of our proposed technique to classical state-of-the-art machine learning methods such as random forest and ensemble models. The empirical results demonstrate the advantages of using neural networks in lieu of an expectation maximization algorithm.
{"title":"A Novel Approach for Synthetic Reduced Nearest-Neighbor Leveraging Neural Networks","authors":"A. Alizadeh, Pooya Tavallali, Vahid Behzadan, A. Ranganath, Mukesh Singhal","doi":"10.1109/ICMLA55696.2022.00138","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00138","url":null,"abstract":"Synthetic Reduced Nearest Neighbor is a nearest neighbor model that is constrained on synthetic samples (i.e., prototypes). The body of work on such models includes proposals for improving the interpretability and optimization of SRNN models using expectation maximization. Motivated by the promise of this paradigm, we propose a novel Expectation Maximization approach for Synthetic Reduced Nearest Neighbors leveraging neural networks. Furthermore, we compare the performance of our proposed technique to classical state-of-the-art machine learning methods such as random forest and ensemble models. The empirical results demonstrate the advantages of using neural networks in lieu of an expectation maximization algorithm.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"352 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115976970","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00178
Hideyuki Masui, D. Romeres, D. Nikovski
Bayesian Optimization has been widely used for black-box optimization. Especially in the field of machine learning, BO has obtained remarkable results in hyperparameters optimization. However, the best hyperparameters depend on the specific task and traditionally the BO algorithm needs to be repeated for each task. On the other hand, the relationship between hyperparameters and objectives has similar tendency among tasks. Therefore, transfer learning is an important technology to accelerate the optimization of novel task by leveraging the knowledge acquired in prior tasks. In this work, we propose a new transfer learning strategy for BO. We use information geometry based principal component analysis (PCA) to extract a low-dimension manifold from a set of Gaussian process (GP) posteriors that models the objective functions of the prior tasks. Then, the low dimensional parameters of this manifold can be optimized to adapt to a new task and set a prior distribution for the objective function of the novel task. Experiments on hyperparameters optimization benchmarks show that our proposed algorithm, called BO-PCA, accelerates the learning of an unseen task (less data are required) while having low computational cost.
{"title":"Transfer Learning for Bayesian Optimization with Principal Component Analysis","authors":"Hideyuki Masui, D. Romeres, D. Nikovski","doi":"10.1109/ICMLA55696.2022.00178","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00178","url":null,"abstract":"Bayesian Optimization has been widely used for black-box optimization. Especially in the field of machine learning, BO has obtained remarkable results in hyperparameters optimization. However, the best hyperparameters depend on the specific task and traditionally the BO algorithm needs to be repeated for each task. On the other hand, the relationship between hyperparameters and objectives has similar tendency among tasks. Therefore, transfer learning is an important technology to accelerate the optimization of novel task by leveraging the knowledge acquired in prior tasks. In this work, we propose a new transfer learning strategy for BO. We use information geometry based principal component analysis (PCA) to extract a low-dimension manifold from a set of Gaussian process (GP) posteriors that models the objective functions of the prior tasks. Then, the low dimensional parameters of this manifold can be optimized to adapt to a new task and set a prior distribution for the objective function of the novel task. Experiments on hyperparameters optimization benchmarks show that our proposed algorithm, called BO-PCA, accelerates the learning of an unseen task (less data are required) while having low computational cost.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133864579","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00225
Justin M. Johnson, T. Khoshgoftaar
There are a variety of data-level and algorithm-level methods available for treating class imbalance. Data-level methods include data sampling strategies that pre-process training data to reduce levels of class imbalance. Algorithm-level methods modify the learning and inference processes to reduce bias towards the majority class. This study evaluates both data-level and algorithm-level methods for class imbalance using a highly imbalanced healthcare fraud data set. We approach the problem from a cost-sensitive learning perspective, and demonstrate how these direct and indirect cost-sensitive methods can be implemented using a common cost matrix. For each method, a wide range of costs are evaluated using three popular ensemble learning algorithms. Initial results show that random undersampling (RUS) and class weighting are both effective ways to improve classification when the default classification threshold is used. Further analysis using the area under the precision-recall curve, however, shows that both RUS and class weighting actually decrease the discriminative power of these learners. Through multiple complementary performance metrics and confidence interval analysis, we find that the best model performance is consistently obtained when RUS and class weighting are not applied, but when output thresholding is used to maximize the confusion matrix instead. Our contributions include various recommendations related to implementing cost-sensitive ensemble learning and effective model evaluation, as well as empirical evidence that contradicts popular beliefs about learning from imbalanced data.
{"title":"Cost-Sensitive Ensemble Learning for Highly Imbalanced Classification","authors":"Justin M. Johnson, T. Khoshgoftaar","doi":"10.1109/ICMLA55696.2022.00225","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00225","url":null,"abstract":"There are a variety of data-level and algorithm-level methods available for treating class imbalance. Data-level methods include data sampling strategies that pre-process training data to reduce levels of class imbalance. Algorithm-level methods modify the learning and inference processes to reduce bias towards the majority class. This study evaluates both data-level and algorithm-level methods for class imbalance using a highly imbalanced healthcare fraud data set. We approach the problem from a cost-sensitive learning perspective, and demonstrate how these direct and indirect cost-sensitive methods can be implemented using a common cost matrix. For each method, a wide range of costs are evaluated using three popular ensemble learning algorithms. Initial results show that random undersampling (RUS) and class weighting are both effective ways to improve classification when the default classification threshold is used. Further analysis using the area under the precision-recall curve, however, shows that both RUS and class weighting actually decrease the discriminative power of these learners. Through multiple complementary performance metrics and confidence interval analysis, we find that the best model performance is consistently obtained when RUS and class weighting are not applied, but when output thresholding is used to maximize the confusion matrix instead. Our contributions include various recommendations related to implementing cost-sensitive ensemble learning and effective model evaluation, as well as empirical evidence that contradicts popular beliefs about learning from imbalanced data.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132581012","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00065
Daniel Bruneß, Matthias Bay, Christian Schulze, Michael Guckert, Mirjam Minor
Automatic classification of documents is a well known problem and can be solved with Machine Learning methods. However, such approaches require large sets of training data which are not always available. Moreover, in data protection sensitive domains, e.g. electronic health records, Machine Learning models often cannot directly be transferred to other environments. We present a transfer learning method which uses ontologies to normalise the feature space of text classifiers. With this we can guarantee that the trained models do not contain any person related data and can therefore be widely reused without raising General Data Protection Regulation (GDPR) issues. Furthermore, we describe a process with which the ontologies can be enriched so that the classifiers can be reused in different contexts with deviating terminology without any additional training of the classifiers. Our transfer learning method follows a combined paradigm of transfer by copy and transfer by enrichment. As proof of concept we apply classifiers trained on hospital medical documents together with appropriately enriched ontologies to medical texts written in colloquial language. The promising results show the potential of our transfer learning approach that respects GDPR requirements and can flexibly be adapted to drifting terminology.
{"title":"An Ontology-based transfer learning method improving classification of medical documents","authors":"Daniel Bruneß, Matthias Bay, Christian Schulze, Michael Guckert, Mirjam Minor","doi":"10.1109/ICMLA55696.2022.00065","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00065","url":null,"abstract":"Automatic classification of documents is a well known problem and can be solved with Machine Learning methods. However, such approaches require large sets of training data which are not always available. Moreover, in data protection sensitive domains, e.g. electronic health records, Machine Learning models often cannot directly be transferred to other environments. We present a transfer learning method which uses ontologies to normalise the feature space of text classifiers. With this we can guarantee that the trained models do not contain any person related data and can therefore be widely reused without raising General Data Protection Regulation (GDPR) issues. Furthermore, we describe a process with which the ontologies can be enriched so that the classifiers can be reused in different contexts with deviating terminology without any additional training of the classifiers. Our transfer learning method follows a combined paradigm of transfer by copy and transfer by enrichment. As proof of concept we apply classifiers trained on hospital medical documents together with appropriately enriched ontologies to medical texts written in colloquial language. The promising results show the potential of our transfer learning approach that respects GDPR requirements and can flexibly be adapted to drifting terminology.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123221144","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00247
YashKiran Prabhu, Naeem Seliya
Stuttering can affect quality of life, resulting in poor social, emotional, and mental health status. Stuttering is diagnosed and managed by speech language pathologists, who are scarce in developing countries. We propose a novel CNN-based Automated Stuttering Identification System (ASIS) to help speech pathologists autonomously diagnose, classify, and log fluency disorders (blocks, prolongations, sound repetitions, word repetitions, and interjections), and monitor patient’s fluency progress over time. A baseline CNN model was created in Tensorflow/Keras and trained and tested using the Sep-28k dataset, an annotated stuttering database of 28,000 3-second clips. We built individual models for each disfluency label and measured accuracy, precision, recall, and F1 measure. The models were built five times, and the averages were taken of each metric. Three different training-validation-test splits were used: 80-10-10, 70-20-10, and 60-20-20. The models performed very well on the public dataset, exceeding the accuracy and F1 measure of other classifiers. The proposed ASIS can help speech pathologists improve the quality of life of stutterers especially in developing countries immensely, and thus it can make a significant difference for millions around the world.
{"title":"A CNN-Based Automated Stuttering Identification System","authors":"YashKiran Prabhu, Naeem Seliya","doi":"10.1109/ICMLA55696.2022.00247","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00247","url":null,"abstract":"Stuttering can affect quality of life, resulting in poor social, emotional, and mental health status. Stuttering is diagnosed and managed by speech language pathologists, who are scarce in developing countries. We propose a novel CNN-based Automated Stuttering Identification System (ASIS) to help speech pathologists autonomously diagnose, classify, and log fluency disorders (blocks, prolongations, sound repetitions, word repetitions, and interjections), and monitor patient’s fluency progress over time. A baseline CNN model was created in Tensorflow/Keras and trained and tested using the Sep-28k dataset, an annotated stuttering database of 28,000 3-second clips. We built individual models for each disfluency label and measured accuracy, precision, recall, and F1 measure. The models were built five times, and the averages were taken of each metric. Three different training-validation-test splits were used: 80-10-10, 70-20-10, and 60-20-20. The models performed very well on the public dataset, exceeding the accuracy and F1 measure of other classifiers. The proposed ASIS can help speech pathologists improve the quality of life of stutterers especially in developing countries immensely, and thus it can make a significant difference for millions around the world.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"488 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126109879","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00013
Jacqueline Höllig, Cedric Kulbach, Steffen Thoma
With the increasing predominance of deep learning methods on time series classification, interpretability becomes essential, especially in high-stake scenarios. Although many approaches to interpretability have been explored for images and tabular data, time series data has been mostly neglected. We approach the problem of interpretability by proposing TSEvo, a model-agnostic multiobjective evolutionary approach to time series counterfactuals incorporating a variety of time series transformation mechanisms to cope with different types and structures of time series. We evaluate our framework on both uni- and multivariate benchmark datasets.
{"title":"TSEvo: Evolutionary Counterfactual Explanations for Time Series Classification","authors":"Jacqueline Höllig, Cedric Kulbach, Steffen Thoma","doi":"10.1109/ICMLA55696.2022.00013","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00013","url":null,"abstract":"With the increasing predominance of deep learning methods on time series classification, interpretability becomes essential, especially in high-stake scenarios. Although many approaches to interpretability have been explored for images and tabular data, time series data has been mostly neglected. We approach the problem of interpretability by proposing TSEvo, a model-agnostic multiobjective evolutionary approach to time series counterfactuals incorporating a variety of time series transformation mechanisms to cope with different types and structures of time series. We evaluate our framework on both uni- and multivariate benchmark datasets.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129867375","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 : 2022-12-01DOI: 10.1109/ICMLA55696.2022.00145
M. F. A. Khan, Hamid Karimi
Fairness in machine learning has become a global concern due to the predominance of ML in automated decision-making systems. In comparison to group fairness, individual fairness, which aspires that similar individuals should be treated similarly, has received limited attention due to some challenges. One major challenge is the availability of a proper metric to evaluate individual fairness, especially for probabilistic classifiers. In this study, we propose a framework PCIndFair to assess the individual fairness of probabilistic classifiers. Unlike current individual fairness measures, our framework considers probability distribution rather than the final classification outcome, which is suitable for capturing the dynamic of probabilistic classifiers, e.g., neural networks. We perform extensive experiments on four standard datasets and discuss the practical benefits of the framework. This study can be helpful for machine learning researchers and practitioners flexibly assess their models' individual fairness. The complete code of the framework is publicly available1.
{"title":"A New Framework to Assess the Individual Fairness of Probabilistic Classifiers","authors":"M. F. A. Khan, Hamid Karimi","doi":"10.1109/ICMLA55696.2022.00145","DOIUrl":"https://doi.org/10.1109/ICMLA55696.2022.00145","url":null,"abstract":"Fairness in machine learning has become a global concern due to the predominance of ML in automated decision-making systems. In comparison to group fairness, individual fairness, which aspires that similar individuals should be treated similarly, has received limited attention due to some challenges. One major challenge is the availability of a proper metric to evaluate individual fairness, especially for probabilistic classifiers. In this study, we propose a framework PCIndFair to assess the individual fairness of probabilistic classifiers. Unlike current individual fairness measures, our framework considers probability distribution rather than the final classification outcome, which is suitable for capturing the dynamic of probabilistic classifiers, e.g., neural networks. We perform extensive experiments on four standard datasets and discuss the practical benefits of the framework. This study can be helpful for machine learning researchers and practitioners flexibly assess their models' individual fairness. The complete code of the framework is publicly available1.","PeriodicalId":128160,"journal":{"name":"2022 21st IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127539726","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：