Filters








80 Hits in 4.0 sec

Optimal transport meets noisy label robust loss and MixUp regularization for domain adaptation [article]

Kilian Fatras, Hiroki Naganuma, Ioannis Mitliagkas
2022 arXiv   pre-print
One strategy to improve these performances is to align the source and target image distributions in an embedded space using optimal transport (OT).  ...  In domain adaptation (DA), one wants to classify unlabeled target images using source labeled images.  ...  Despite these weaknesses, optimal transport and its minibatch approximation have found numerous applications to compare probability distributions from different domains as in domain adaptation.  ... 
arXiv:2206.11180v1 fatcat:3fjrwyaidrfjhkqm6csw523n5q

Improving Mini-batch Optimal Transport via Partial Transportation [article]

Khai Nguyen, Dang Nguyen, The-Anh Vu-Le, Tung Pham, Nhat Ho
2022 arXiv   pre-print
Finally, we carry out extensive experiments on various applications such as deep domain adaptation, partial domain adaptation, deep generative model, color transfer, and gradient flow to demonstrate the  ...  Mini-batch optimal transport (m-OT) has been widely used recently to deal with the memory issue of OT in large-scale applications.  ...  Unbalanced minibatch optimal transport; applications to domain adaptation. In Meila, M. and Zhang, T. (eds.), Figalli, A. The optimal partial transport problem.  ... 
arXiv:2108.09645v4 fatcat:zlaypaozl5hqtp5rng2vrc2zky

Unsupervised Domain Adaptation for LiDAR Panoptic Segmentation [article]

Borna Bešić, Nikhil Gosala, Daniele Cattaneo, Abhinav Valada
2021 arXiv   pre-print
Unsupervised Domain Adaptation (UDA) techniques are thus essential to fill this domain gap and retain the performance of models on new sensor setups without the need for additional data labeling.  ...  While data-based adaptations reduce the domain gap by processing the raw LiDAR scans to resemble the scans in the target domain, model-based techniques guide the network in extracting features that are  ...  Specifically, we solve the discrete unbalanced optimal transport between the source and target feature minibatches using a recent extension to the Sinkhorn algorithm [24] .  ... 
arXiv:2109.15286v1 fatcat:ctn7spzbejcpldzrd7sbwhnm44

Embedding Signals on Knowledge Graphs with Unbalanced Diffusion Earth Mover's Distance [article]

Alexander Tong and Guillaume Huguet and Dennis Shung and Amine Natik and Manik Kuchroo and Guillaume Lajoie and Guy Wolf and Smita Krishnaswamy
2022 arXiv   pre-print
Typically, EMD is computed by optimizing over the cost of transporting one probability distribution to another over an underlying metric space.  ...  In modern relational machine learning it is common to encounter large graphs that arise via interactions or similarities between observations in many domains.  ...  However, as we show here, the construction can be adapted to consider unbalanced transport, which is essentially based on the idea that a more faithful earth mover's distance is given by a transport in  ... 
arXiv:2107.12334v2 fatcat:ey4acpcqqzg5rolceuexmozs2q

Limited Data Rolling Bearing Fault Diagnosis with Few-shot Learning

Ansi Zhang, Shaobo Li, Yuxin Cui, Wanli Yang, Rongzhi Dong, Jianjun Hu
2019 IEEE Access  
We use domain adaptation to simulate new work conditions.  ...  TABLE 5 . 5 Scenario setting for domain adaptation. TABLE 6 . 6 Complexity of tree datasets. TABLE 7 . 7 New scenario setting for domain adaptation.  ...  He was a Postdoctoral Fellow in bioinformatics with Purdue University and the University of Southern California, from 2004 to 2007.  ... 
doi:10.1109/access.2019.2934233 fatcat:sxhehpp5hnb57c7n5tkteybvbm

Semi-supervised Optimal Transport with Self-paced Ensemble for Cross-hospital Sepsis Early Detection [article]

Ruiqing Ding, Yu Zhou, Jie Xu, Yan Xie, Qiqiang Liang, He Ren, Yixuan Wang, Yanlin Chen, Leye Wang, Man Huang
2021 arXiv   pre-print
the semi-supervised domain adaptation based on optimal transport theory with self-paced under-sampling to avoid a negative transfer possibly caused by covariate shift and class imbalance.  ...  In this paper, we propose a semi-supervised optimal transport with self-paced ensemble framework for Sepsis early detection, called SPSSOT, to transfer knowledge from the other that has rich labeled data  ...  More specifically, in semi-supervised domain adaptation with optimal transport, we devise a label adaptive optimal transport strategy to achieve the precise-pairwise optimal transport and an intra-domain  ... 
arXiv:2106.10352v1 fatcat:wwmoxns35jbdzpqaq77rgoh6qe

Metric Learning-enhanced Optimal Transport for Biochemical Regression Domain Adaptation [article]

Fang Wu, Nicolas Courty, Zhang Qiang, jiyu Cui, Ziqing Li
2022 arXiv   pre-print
To meet this challenge, researchers have used optimal transport (OT) to perform representation alignment between the source and target domains.  ...  To exploit continuous labels, we propose novel metrics to measure domain distances and introduce a posterior variance regularizer on the transport plan.  ...  Optimal Transport for Regressions Unsupervised Domain Adaptation UDA is common in biochemistry.  ... 
arXiv:2202.06208v2 fatcat:yodnvw4ldbg3boknp77io5ttpa

On Transportation of Mini-batches: A Hierarchical Approach [article]

Khai Nguyen, Dang Nguyen, Quoc Nguyen, Tung Pham, Hung Bui, Dinh Phung, Trung Le, Nhat Ho
2022 arXiv   pre-print
Finally, we carry out experiments on various applications including deep generative models, deep domain adaptation, approximate Bayesian computation, color transfer, and gradient flow to show that the  ...  To address these problems, we propose a novel mini-batch scheme for optimal transport, named Batch of Mini-batches Optimal Transport (BoMb-OT), that finds the optimal coupling between mini-batches and  ...  C Applications and BoMb-OT Algorithms In this section, we collect the details of applications that mini-batch optimal transport is used in practice including deep generative models, deep domain adaptation  ... 
arXiv:2102.05912v5 fatcat:6kpvqm2iirdzlgusdpnlddk7ki

Collective Traffic Forecasting [chapter]

Marco Lippi, Matteo Bertini, Paolo Frasconi
2010 Lecture Notes in Computer Science  
This is a highly relational task along the spatial and the temporal dimensions and we advocate the application of statistical relational learning techniques.  ...  Traffic forecasting has recently become a crucial task in the area of intelligent transportation systems, and in particular in the development of traffic management and control.  ...  modules are used to adapt system variables and maintain optimal performances.  ... 
doi:10.1007/978-3-642-15883-4_17 fatcat:y5eugejyajdkjlgwu463leu7t4

Distributional Sliced-Wasserstein and Applications to Generative Modeling [article]

Khai Nguyen and Nhat Ho and Tung Pham and Hung Bui
2020 arXiv   pre-print
Finally, we conduct extensive experiments with large-scale datasets to demonstrate the favorable performances of the proposed distances over the previous sliced-based distances in generative modeling applications  ...  In order to account for these weaknesses, we propose a novel distance, named Distributional Sliced-Wasserstein distance (DSW), that finds an optimal distribution over projections that can balance between  ...  Joint distribution optimal transportation for domain adaptation. In Advances in Neural Information Processing Systems, [17] V. Dumoulin, I. Belghazi, B. Poole, O. Mastropietro, A. Lamb, M.  ... 
arXiv:2002.07367v2 fatcat:tlmpjroke5aifiqy34oqrfmxo4

SDN-Based Architecture for Transport and Application Layer DDoS Attack Detection by Using Machine and Deep Learning

Noe M. Yungaicela-Naula, Cesar Vargas-Rosales, Jesus Arturo Perez-Diaz
2021 IEEE Access  
ACKNOWLEDGMENT The authors would like to thank Maritza Rosales Hernández, Fátima Sánchez Suárez, and Martín Helmut Domínguez Álvarez, for their assistance in training and testing the ML/DL models using  ...  In addition, most works studied transport or application layer attacks separately. We propose to evaluate different ML/DL techniques on both transport and application layer DoS/DDoS attacks.  ...  In this work, we designed an intelligent solution to detect transport and application layer DoS/DDoS attacks.  ... 
doi:10.1109/access.2021.3101650 fatcat:4x2clm3anvcijolphwlnaqzacq

On Target Shift in Adversarial Domain Adaptation [article]

Yitong Li, Michael Murias, Samantha Major, Geraldine Dawson, David E. Carlson
2019 arXiv   pre-print
We extend this framework to handle multiple domains by developing a scheme to upweight source domains most similar to the target domain.  ...  In this work, we propose a method called Domain Adversarial nets for Target Shift (DATS) to address label shift while learning a domain invariant representation.  ...  Recently, optimal transport has been used to analyze the problem of label shift in domain adaptation [26] , but did not consider learning a feature extractor in conjunction with their framework.  ... 
arXiv:1903.06336v1 fatcat:jpd63yc3jfa5hdfg46mrfqd5ba

Connecting GANs, MFGs, and OT [article]

Haoyang Cao, Xin Guo, Mathieu Laurière
2021 arXiv   pre-print
the optimal transport cost indexed by the generator from the known latent distribution to the unknown true distribution of data.  ...  More specifically, from the game theoretical perspective, GANs are interpreted as MFGs under Pareto Optimality criterion or mean-field controls; from the optimal transport perspective, GANs are to minimize  ...  An unbalanced optimal transport problem is solved using GANs in [58] .  ... 
arXiv:2002.04112v4 fatcat:yg7b3baepvfdte6isbhnuzsevm

Bearing Fault Detection and Diagnosis Using Case Western Reserve University Dataset With Deep Learning Approaches: A Review

Dhiraj Neupane, Jongwon Seok
2020 IEEE Access  
With the increment in the use of smart machinery, the faults in the machinery equipment are expected to increase.  ...  This paper, we believe, can be of good help for future researchers to start their work on machinery fault detection and diagnosis using the CWRU dataset.  ...  ACKNOWLEDGMENT The authors thank Case Western Reserve University for providing free access to the bearing vibration experimental data from their website.  ... 
doi:10.1109/access.2020.2990528 fatcat:252hcj5d5bftxedititg2ya7sm

Anomaly Detection Based on Convex Analysis: A Survey

Tong Wang, Mengsi Cai, Xiao Ouyang, Ziqiang Cao, Tie Cai, Xu Tan, Xin Lu
2022 Frontiers in Physics  
, and mathematical optimization for modeling.  ...  Convex analysis (CA) is one of the fundamental methods used in anomaly detection, which contributes to the robust approximation of algebra and geometry, efficient computation to a unique global solution  ...  few and two classes are extremely unbalanced, the generalization ability of machine learning methods, especially the gradient descent-based model, should be strengthened to be more suitable and applicable  ... 
doi:10.3389/fphy.2022.873848 fatcat:ooxtghts5ffoxmu2qx743ij3eq
« Previous Showing results 1 — 15 out of 80 results