Robust feature learning for adversarial defense via hierarchical feature alignment

Zhang, Xiaoqin; Wang, Jinxin; Wang, Tao; Jiang, Runhua; Xu, Jiawei; Zhao, Li

doi:10.1016/j.ins.2020.12.042

Cited by 63 publications

(21 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…e importance of the criteria has been determined based on studies of theoretical principles. One future direction in this aspect can be the application of machine learning models [42][43][44][45][46][47][48], decision-making approaches [8,49,50], enhanced deep learning models [51][52][53][54][55], machine learning-based tools [56,57], deep learning methods [58], optimized prediction tasks [59][60][61][62][63], optimization algorithms [64][65][66][67][68], and intelligent systems and hybrid feature selection methods for weighting the importance of the criteria.…”

Section: Research Findingsmentioning

confidence: 99%

Process-Based Improvement of Urban Metabolism in Optimizing the Development Cycle of the Small City Using MIA Method

Delivandani

Rajabi

Kermani

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

In order to maintain the health and stability of the urban ecosystem, humans must undermine the negative effects of improper land use in cities by planning to reduce entropy and regulate urban metabolism, material, and energy cycles, and considering resource capacity when providing the needs of the urban ecosystem population. Subsequently, the general purpose of this study is to explain the process-based pattern of healthy urban metabolism in the development cycle of a small city. The purpose of studying the urban metabolism, which is an integral part of the government from the environment, is to prepare reports and take measures indicating city sustainability. Indeed, the urban metabolism incorporates the relevant information on energy saving, material cycle, and management of waste and infrastructure in the urban systems. The present study employs a descriptive-analytical method. Accordingly, the metabolic impact assessment (MIA) method was utilized to analyze the data and achieve the study results. According to the assessment results in the input part of the urban metabolism process, the water and energy criteria are closer to the ideal status with 64% and 40%, respectively. In total, the performance of the input process is equal to 52%, and after that, the air quality, materials, and output sectors have a performance equal to 35, 31, and 33%, respectively. Moreover, land cover and transport have a performance of 14% and 65%, respectively, revealing that they are in a desirable condition. The above results based on mathematical optimization illustrate that there is no balance between the input and output of the urban metabolism model in the study area, and the main problems are evident in the output sectors and particularly in the recycling of materials and water.

show abstract

Section: Research Findingsmentioning

confidence: 99%

Process-Based Improvement of Urban Metabolism in Optimizing the Development Cycle of the Small City Using MIA Method

Delivandani

Rajabi

Kermani

2021

Mathematical Problems in Engineering

View full text Add to dashboard Cite

show abstract

“…PixelMask [2], adversarial defense via hierarchical feature alignment [42], etc.). The advantage of using white-box defense is it can provide higher performance for known attacks than blackbox defense.…”

Section: Defensesmentioning

confidence: 99%

Attack-less adversarial training for a robust adversarial defense

Lee

Kang

2021

Appl Intell

View full text Add to dashboard Cite

Adversarial examples have proved efficacious in fooling deep neural networks recently. Many researchers have studied this issue of adversarial examples by evaluating neural networks against their attack techniques and increasing the robustness of neural networks with their defense techniques. To the best of our knowledge, adversarial training is one of the most effective defense techniques against the adversarial examples. However, the method is not able to cope with new attacks because it requires attack techniques in the training phase. In this paper, we propose a novel defense technique, Attack-Less Adversarial Training (ALAT) method, which is independent from any attack techniques, thereby is useful in preventing future attacks. Specifically, ALAT regenerates every pixel of an image into different pixel value, which commonly eliminates the majority of the adversarial noises in the adversarial example. This pixel regeneration is useful in defense because the adversarial noises are the core problem that make the neural networks produce high misclassification rate. Our experiment results with several benchmark datasets show that our method not only relieves over-fitting issue during the training of neural networks with a large number of epochs, but also boosts the robustness of the neural network.

show abstract

“…Such types of images play a critical role in diagnosis of the COVID-19 side effects. Furthermore, advanced machine learning techniques have shown their great significance progress in addressing many practical applications such as recovering low-rank tensors [12] , feature selection [13] , [14] , [15] , information retrieval [16] , [17] , [18] , [19] , [20] , and computer vision [21] , [22] , [22] , [23] , [23] , [24] , [24] , [25] . By utilizing the excellent benefits of an advanced machine learning strategy called deep learning with analysis of X-ray and CT images can benefit health care providers with treatment assists for COVID-19, which leads in contributing to higher insights into the virus progresses.…”

Section: Introductionmentioning

confidence: 99%

An oppositional-Cauchy based GSK evolutionary algorithm with a novel deep ensemble reinforcement learning strategy for COVID-19 diagnosis

Jalali

Ahmadian

et al. 2021

Applied Soft Computing

View full text Add to dashboard Cite

A novel coronavirus (COVID-19) has globally attracted attention as a severe respiratory condition. The epidemic has been first tracked in Wuhan, China, and has progressively been expanded in the entire world. The growing expansion of COVID-19 around the globe has made X-ray images crucial for accelerated diagnostics. Therefore, an effective computerized system must be established as a matter of urgency, to facilitate health care professionals in recognizing X-ray images from COVID-19 patients. In this work, we design a novel artificial intelligent-based automated X-ray image analysis framework based on an ensemble of deep optimized convolutional neural networks (CNNs) in order to distinguish coronavirus patients from non-patients. By developing a modified version of gaining–sharing knowledge (GSK) optimization algorithm using the Opposition-based learning (OBL) and Cauchy mutation operators, the architectures of the deployed deep CNNs are optimized automatically without performing the general trial and error procedures. After obtaining the optimized CNNs, it is also very critical to identify how to decrease the number of ensemble deep CNN classifiers to ensure the classification effectiveness. To this end, a selective ensemble approach is proposed for COVID-19 X-ray based image classification using a deep Q network that combines reinforcement learning (RL) with the optimized CNNs. This approach increases the model performance in particular and therefore decreases the ensemble size of classifiers. The experimental results show that the proposed deep RL optimized ensemble approach has an excellent performance over two popular X-ray image based COVID-19 datasets. Our proposed advanced algorithm can accurately identify the COVID-19 patients from the normal individuals with a significant accuracy of 0.991441, precision of 0.993568, recall (sensitivity) of 0.981445, F-measure of 0.989666 and AUC of 0.990337 for Kaggle dataset as well as an excellent accuracy of 0.987742, precision of 0.984334, recall (sensitivity) of 0.989123, F-measure of 0.984939 and AUC of 0.988466 for Mendely dataset.

show abstract

Robust feature learning for adversarial defense via hierarchical feature alignment

Cited by 63 publications

References 12 publications

Process-Based Improvement of Urban Metabolism in Optimizing the Development Cycle of the Small City Using MIA Method

Process-Based Improvement of Urban Metabolism in Optimizing the Development Cycle of the Small City Using MIA Method

Attack-less adversarial training for a robust adversarial defense

An oppositional-Cauchy based GSK evolutionary algorithm with a novel deep ensemble reinforcement learning strategy for COVID-19 diagnosis

Contact Info

Product

Resources

About