检索结果-内蒙古大学图书馆

Comparing Feature Importance and Rule Extraction for Interpretability on Text Data

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Lopardo, Gianluigi Garreau, Damien Université Côte d’Azur Inria CNRS LJAD France

Complex machine learning algorithms are used more and more often in critical tasks involving text data, leading to the development of interpretability methods. Among local methods, two families have emerged: those computing importance scores for each feature and those extracting simple logical rules. In this paper we show that using different methods can lead to unexpectedly different explanations, even when applied to simple models for which we would expect qualitative coincidence. To quantify this effect, we propose a new approach to compare explanations produced by different methods. Copyright © 2022, The Authors. All rights reserved.

关键词： learning algorithms

Deep Active Ensemble Sampling For Image Classification

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Mohamadi, Salman Doretto, Gianfranco Adjeroh, Donald A. West Virginia University MorgantownWV United States

Conventional active learning (AL) frameworks aim to reduce the cost of data annotation by actively requesting the labeling for the most informative data points. However, introducing AL to data hungry deep learning algorithms has been a challenge. Some proposed approaches include uncertainty-based techniques, geometric methods, implicit combination of uncertainty-based and geometric approaches, and more recently, frameworks based on semi/self supervised techniques. In this paper, we address two specific problems in this area. The first is the need for efficient exploitation/exploration trade-off in sample selection in AL. For this, we present an innovative integration of recent progress in both uncertainty-based and geometric frameworks to enable an efficient exploration/exploitation trade-off in sample selection strategy. To this end, we build on a computationally efficient approximate of Thompson sampling with key changes as a posterior estimator for uncertainty representation. Our framework provides two advantages: (1) accurate posterior estimation, and (2) tune-able trade-off between computational overhead and higher accuracy. The second problem is the need for improved training protocols in deep AL. For this, we use ideas from semi/self supervised learning to propose a general approach that is independent of the specific AL technique being used. Taken these together, our framework shows a significant improvement over the state-of-the-art, with results that are comparable to the performance of supervised-learning under the same setting. We show empirical results of our framework, and comparative performance with the state-of-the-art on four datasets, namely, MNIST, CIFAR10, CIFAR100 and ImageNet to establish a new baseline in two different settings. © 2022, CC BY.

关键词： learning algorithms

Quasi-Balanced Self-Training on Noise-Aware Synthesis of Object Point Clouds for Closing Domain Gap

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Chen, Yongwei Wang, Zihao Zou, Longkun Chen, Ke Jia, Kui South China University of Technology China DexForce Co. Ltd China Peng Cheng Laboratory China

Semantic analyses of object point clouds are largely driven by releasing of benchmarking datasets, including synthetic ones whose instances are sampled from object CAD models. However, learning from synthetic data may not generalize to practical scenarios, where point clouds are typically incomplete, non-uniformly distributed, and noisy. Such a challenge of Simulation-to-Reality (Sim2Real) domain gap could be mitigated via learning algorithms of domain adaptation;however, we argue that generation of synthetic point clouds via more physically realistic rendering is a powerful alternative, as systematic non-uniform noise patterns can be captured. To this end, we propose an integrated scheme consisting of physically realistic synthesis of object point clouds via rendering stereo images via projection of speckle patterns onto CAD models and a novel quasi-balanced self-training designed for more balanced data distribution by sparsity-driven selection of pseudo labeled samples for long tailed classes. Experiment results can verify the effectiveness of our method as well as both of its modules for unsupervised domain adaptation on point cloud classification, achieving the state-of-the-art performance. Source codes and the SpeckleNet synthetic dataset are available at https://***/Gorilla-Lab-SCUT/QS3. Copyright © 2022, The Authors. All rights reserved.

关键词： learning algorithms

Analysis of Intelligent Approaches for Discovery and Management of Knowledge: A Review

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Ahamad, Rayees Mishra, Kamta Nath Department of Computer Science and Engineering India Birla Institute of Technology Mesra India

Knowledge Discovery and Management (KDM) is a process and approach for creating, discovery, capturing, organizing, refining, presenting, and providing data, information, and knowledge for a specific goal. Knowledge Management (KM) and Artificial Intelligence (AI) at their core are about the knowledge. AI provides the mechanism that allows a machine to obtain, acquire, process, and use information in order to execute tasks, as well as reveal or unlock knowledge that may be transmitted to people for improved strategic decision-making. There are many conventional methods used in the process of KDM. But, the use of intelligent approaches in this process can further improve the efficiency in the sense of time and accuracy. Intelligent techniques especially soft computing approaches have the ability to learn in any environment with the help of logics, reasoning, and other computing abilities. Intelligent techniques may be categorized as learning algorithms (Supervised, Unsupervised, and Reinforcement), Logic & Rule-Based algorithms (Fuzzy Logic, Bayesian Network, and CBR-RBR), Nature-inspired algorithms (Genetic algorithm, Particle Swarm Optimization, and Ant Colony Optimization), and hybrid approaches (combination of these algorithms). The main aim of existing intelligent techniques is to solve the day-to-day problems of our rural and smart digital societies. In this paper, the authors have studied many intelligent computing methods (ICMs) that are related to specific problems and provide the correct and reasonable solutions in the form of knowledge. Single ICM and combined ICMs are employed to solve the domain-specific problems are also discussed and analyzed in this study. It is observed from the results that the combined ICMs have better efficiency than the single ICM. Finally, the authors have presented the analysis and comparison of ICMs based on their application domain, parameters, methods/algorithms, efficiency and acceptable outcomes. Further, the authors have

关键词： learning algorithms

Pneumonia Detection in Chest X-Rays using Neural Networks

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Darapaneni, Narayana Ranjan, Ashish Bright, Dany Trivedi, Devendra Kumar, Ketul Kumar, Vivek Paduri, Anwesh Reddy Northwestern University Great Learning United States Great Learning Bangalore India

With the advancement in AI, deep learning techniques are widely used to design robust classification models in several areas such as medical diagnosis tasks in which it achieves good performance. In this paper, we have proposed the CNN model (Convolutional Neural Network) for the classification of Chest X-ray images for Radiological Society of North America Pneumonia (RSNA) datasets. The study also tries to achieve the same RSNA benchmark results using the limited computational resources by trying out various approaches to the methodologies that have been implemented in recent years. The proposed method is based on a non-complex CNN and the use of transfer learning algorithms like Xception, InceptionV3/V4, EfficientNetB7. Along with this, the study also tries to achieve the same RSNA benchmark results using the limited computational resources by trying out various approaches to the methodologies that have been implemented in recent years. The RSNA benchmark MAP score is 0.25, but using the Mask RCNN model on a stratified sample of 3017 along with image augmentation gave a MAP score of 0.15. Meanwhile, the YoloV3 without any hyperparameter tuning gave the MAP score of 0.32 but still, the loss keeps decreasing. Running the model for a greater number of iterations can give better results. © 2022, CC BY-NC-ND.

关键词： learning algorithms

Exact Privacy Guarantees for Markov Chain Implementations of the Exponential Mechanism with Artificial Atoms

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Seeman, Jeremy Reimherr, Matthew Slavković, Aleksandra Department of Statistics Penn State University State CollegePA16803 United States

Implementations of the exponential mechanism in differential privacy often require sampling from intractable distributions. When approximate procedures like Markov chain Monte Carlo (MCMC) are used, the end result incurs costs to both privacy and accuracy. Existing work has examined these effects asymptotically, but implementable finite sample results are needed in practice so that users can specify privacy budgets in advance and implement samplers with exact privacy guarantees. In this paper, we use tools from ergodic theory and perfect simulation to design exact finite runtime sampling algorithms for the exponential mechanism by introducing an intermediate modified target distribution using artificial atoms. We propose an additional modification of this sampling algorithm that maintains its ϵ-DP guarantee and has improved runtime at the cost of some utility. We then compare these methods in scenarios where we can explicitly calculate a δ cost (as in (ϵ, δ)-DP) incurred when using standard MCMC techniques. Much as there is a well known trade-off between privacy and utility, we demonstrate that there is also a trade-off between privacy guarantees and runtime. © 2022, CC BY.

关键词： learning algorithms

An Empirical Study on Distribution Shift Robustness from the Perspective of Pre-Training and Data Augmentation

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Liu, Ziquan Xu, Yi Xu, Yuanhong Qian, Qi Li, Hao Jin, Rong Ji, Xiangyang Chan, Antoni B. Department of Computer Science City University of Hong Kong Hong Kong School of Artificial Intelligence Dalian University of Technology China DAMO Academy Alibaba Group China Department of Automation Tsinghua University China

The performance of machine learning models under distribution shift has been the focus of the community in recent years. Most of current methods have been proposed to improve the robustness to distribution shift from the algorithmic perspective, i.e., designing better training algorithms to help the generalization in shifted test distributions. This paper studies the distribution shift problem from the perspective of pre-training and data augmentation, two important factors in the practice of deep learning that have not been systematically investigated by existing work. By evaluating seven pre-trained models, including ResNets [1] and ViT's [2] with self-supervision and supervision mode, on five important distributionshift datasets, from WILDS [3] and DomainBed [4] benchmarks, with five different learning algorithms, we provide the first comprehensive empirical study focusing on pre-training and data augmentation. With our empirical result obtained from 1,330 models, we provide the following main observations: 1) ERM combined with data augmentation can achieve state-of-the-art performance if we choose a proper pre-trained model respecting the data property;2) specialized algorithms further improve the robustness on top of ERM when handling a specific type of distribution shift, e.g., GroupDRO [5] for spurious correlation and CORAL [6] for large-scale out-of-distribution data;3) Comparing different pre-training modes, architectures and data sizes, we provide novel observations about pre-training on distribution shift, which sheds light on designing or selecting pre-training strategy for different kinds of distribution shifts. In summary, our empirical study provides a comprehensive baseline for a wide range of pre-training models fine-tuned with data augmentation, which potentially inspires research exploiting the power of pre-training and data augmentation in the future of distribution shift study. © 2022, CC BY.

关键词： learning algorithms

MATRIX MULTIPLICATIVE WEIGHTS UPDATES IN QUANTUM ZERO-SUM GAMES: CONSERVATION LAWS & RECURRENCE

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Jain, Rahul Piliouras, Georgios Sim, Ryann Centre for Quantum Technologies Department of Computer Science National University of Singapore MajuLab UMI3654 Singapore Engineering Systems and Design Singapore University of Technology and Design Singapore

Recent advances in quantum computing and in particular, the introduction of quantum GANs, have led to increased interest in quantum zero-sum game theory, extending the scope of learning algorithms for classical games into the quantum realm. In this paper, we focus on learning in quantum zero-sum games under Matrix Multiplicative Weights Update (a generalization of the multiplicative weights update method) and its continuous analogue, Quantum Replicator Dynamics. When each player selects their state according to quantum replicator dynamics, we show that the system exhibits conservation laws in a quantum-information theoretic sense. Moreover, we show that the system exhibits Poincaré recurrence, meaning that almost all orbits return arbitrarily close to their initial conditions infinitely often. Our analysis generalizes previous results in the case of classical games [43, 49]. Copyright © 2022, The Authors. All rights reserved.

关键词： learning algorithms

On Error and Compression Rates for Prototype Rules

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Kerem, Omer Weiss, Roi Ben-Gurion University of the Negev Ariel University

We study the close interplay between error and compression in the non-parametric multiclass classification setting in terms of prototype learning rules. We focus in particular on a recently proposed compression-based learning rule termed OptiNet (Kontorovich, Sabato, and Urner 2016;Kontorovich, Sabato, and Weiss 2017;Hanneke et al. 2021). Beyond its computational merits, this rule has been recently shown to be universally consistent in any metric instance space that admits a universally consistent rule—the first learning algorithm known to enjoy this property. However, its error and compression rates have been left open. Here we derive such rates in the case where instances reside in Euclidean space under commonly posed smoothness and tail conditions on the data distribution. We first show that OptiNet achieves non-trivial compression rates while enjoying near minimax-optimal error rates. We then proceed to study a novel general compression scheme for further compressing prototype rules that locally adapts to the noise level without sacrificing accuracy. Applying it to OptiNet, we show that under a geometric margin condition, further gain in the compression rate is achieved. Experimental results comparing the performance of the various methods are presented. © 2022, CC BY.

关键词： learning algorithms