complex models
Recently Published Documents


TOTAL DOCUMENTS

1105
(FIVE YEARS 387)

H-INDEX

52
(FIVE YEARS 8)

2022 ◽  
Vol 2022 (1) ◽  
Author(s):  
Jing Lin ◽  
Laurent L. Njilla ◽  
Kaiqi Xiong

AbstractDeep neural networks (DNNs) are widely used to handle many difficult tasks, such as image classification and malware detection, and achieve outstanding performance. However, recent studies on adversarial examples, which have maliciously undetectable perturbations added to their original samples that are indistinguishable by human eyes but mislead the machine learning approaches, show that machine learning models are vulnerable to security attacks. Though various adversarial retraining techniques have been developed in the past few years, none of them is scalable. In this paper, we propose a new iterative adversarial retraining approach to robustify the model and to reduce the effectiveness of adversarial inputs on DNN models. The proposed method retrains the model with both Gaussian noise augmentation and adversarial generation techniques for better generalization. Furthermore, the ensemble model is utilized during the testing phase in order to increase the robust test accuracy. The results from our extensive experiments demonstrate that the proposed approach increases the robustness of the DNN model against various adversarial attacks, specifically, fast gradient sign attack, Carlini and Wagner (C&W) attack, Projected Gradient Descent (PGD) attack, and DeepFool attack. To be precise, the robust classifier obtained by our proposed approach can maintain a performance accuracy of 99% on average on the standard test set. Moreover, we empirically evaluate the runtime of two of the most effective adversarial attacks, i.e., C&W attack and BIM attack, to find that the C&W attack can utilize GPU for faster adversarial example generation than the BIM attack can. For this reason, we further develop a parallel implementation of the proposed approach. This parallel implementation makes the proposed approach scalable for large datasets and complex models.


Sensors ◽  
2022 ◽  
Vol 22 (2) ◽  
pp. 551
Author(s):  
Chih-Wei Lin ◽  
Xiuping Huang ◽  
Mengxiang Lin ◽  
Sidi Hong

Precipitation intensity estimation is a critical issue in the analysis of weather conditions. Most existing approaches focus on building complex models to extract rain streaks. However, an efficient approach to estimate the precipitation intensity from surveillance cameras is still challenging. This study proposes a convolutional neural network known as the signal filtering convolutional neural network (SF-CNN) to handle precipitation intensity using surveillance-based images. The SF-CNN has two main blocks, the signal filtering block (SF block) and the gradually decreasing dimension block (GDD block), to extract features for the precipitation intensity estimation. The SF block with the filtering operation is constructed in different parts of the SF-CNN to remove the noise from the features containing rain streak information. The GDD block continuously takes the pair of the convolutional operation with the activation function to reduce the dimension of features. Our main contributions are (1) an SF block considering the signal filtering process and effectively removing the useless signals and (2) a procedure of gradually decreasing the dimension of the feature able to learn and reserve the information of features. Experiments on the self-collected dataset, consisting of 9394 raining images with six precipitation intensity levels, demonstrate the proposed approach’s effectiveness against the popular convolutional neural networks. To the best of our knowledge, the self-collected dataset is the largest dataset for monitoring infrared images of precipitation intensity.


2022 ◽  
pp. 004912412110675
Author(s):  
Soojin Park ◽  
Xu Qin ◽  
Chioun Lee

In the field of disparities research, there has been growing interest in developing a counterfactual-based decomposition analysis to identify underlying mediating mechanisms that help reduce disparities in populations. Despite rapid development in the area, most prior studies have been limited to regression-based methods, undermining the possibility of addressing complex models with multiple mediators and/or heterogeneous effects. We propose a novel estimation method that effectively addresses complex models. Moreover, we develop a sensitivity analysis for possible violations of an identification assumption. The proposed method and sensitivity analysis are demonstrated with data from the Midlife Development in the US study to investigate the degree to which disparities in cardiovascular health at the intersection of race and gender would be reduced if the distributions of education and perceived discrimination were the same across intersectional groups.


Energies ◽  
2022 ◽  
Vol 15 (2) ◽  
pp. 487
Author(s):  
Bilin Shao ◽  
Yichuan Yan ◽  
Huibin Zeng

Accurate short-term load forecasting can ensure the safe operation of the grid. Decomposing load data into smooth components by decomposition algorithms is a common approach to address data volatility. However, each component of the decomposition must be modeled separately for prediction, which leads to overly complex models. To solve this problem, a VMD-WSLSTM load prediction model based on Shapley values is proposed in this paper. First, the Shapley value is used to select the optimal set of special features, and then the VMD decomposition method is used to decompose the original load into several smooth components. Finally, WSLSTM is used to predict each component. Unlike the traditional LSTM model, WSLSTM can simplify the prediction model and extract common features among the components by sharing the parameters among the components. In order to verify the effectiveness of the proposed model, several control groups were used for experiments. The results show that the proposed method has higher prediction accuracy and training speed compared with traditional prediction methods.


2022 ◽  
Vol 15 (1) ◽  
pp. 29
Author(s):  
Rainer Baule ◽  
Philip Rosenthal

Hedging down-and-out puts (and up-and-out calls), where the maximum payoff is reached just before a barrier is hit that would render the claim worthless afterwards, is challenging. All hedging methods potentially lead to large errors when the underlying is already close to the barrier and the hedge portfolio can only be adjusted in discrete time intervals. In this paper, we analyze this hedging situation, especially the case of overnight trading gaps. We show how a position in a short-term vanilla call option can be used for efficient hedging. Using a mean-variance hedging approach, we calculate optimal hedge ratios for both the underlying and call options as hedge instruments. We derive semi-analytical formulas for optimal hedge ratios in a Black–Scholes setting for continuous trading (as a benchmark) and in the case of trading gaps. For more complex models, we show in a numerical study that the semi-analytical formulas can be used as a sufficient approximation, even when stochastic volatility and jumps are present.


Queue ◽  
2021 ◽  
Vol 19 (6) ◽  
pp. 28-56
Author(s):  
Valerie Chen ◽  
Jeffrey Li ◽  
Joon Sik Kim ◽  
Gregory Plumb ◽  
Ameet Talwalkar

The emergence of machine learning as a society-changing technology in the past decade has triggered concerns about people's inability to understand the reasoning of increasingly complex models. The field of IML (interpretable machine learning) grew out of these concerns, with the goal of empowering various stakeholders to tackle use cases, such as building trust in models, performing model debugging, and generally informing real human decision-making.


Author(s):  
A.М. Заяц ◽  
С.П. Хабаров

Предложен подход к разработке в среде OMNeT++ INET простейшей имитационной модели инфраструктурного режима функционирования Wi-Fi сети, который позволяет проводить подробный анализ функционирования таких сетей, а также строить и анализировать временные диаграммы взаимодействия всех элементов сети. Разработанную модель можно использовать как базовую для формирования более сложных моделей с произвольным числом мобильных клиентов, позволяя определять необходимое количество точек доступа и мест их размещения для обеспечения полноценного покрытия зоны мониторинга лесной территории. An approach to the development in the OMNeT ++ INET environment of the simplest simulation model of the infrastructure mode of Wi-Fi network operation is proposed, which allows a detailed analysis of the functioning of such networks, as well as to build and analyze the time diagram of the interaction of all network elements. The developed model can be used as a base for the formation of more complex models with an arbitrary number of mobile clients, allowing you to determine the required number of access points and their locations to ensure full coverage of the monitoring area of the forest area.


2021 ◽  
Author(s):  
Khoi D Nguyen ◽  
Madhusudhan Venkadesan

Muscle rheology, or the characterization of a muscle's response to external mechanical perturbations, is crucial to an animal's motor control and locomotive abilities. How the rheology emerges from the ensemble dynamics of microscopic actomyosin crossbridges known to underlie muscle forces is however a longstanding question. Classical descriptions in terms of force-length and force-velocity relationships capture only part of the rheology, namely under steady but not dynamical conditions. Although much is known about the actomyosin machinery, current mathematical models that describe the behavior of a population or an ensemble of crossbridges are plagued by an excess of parameters and computational complexity that limits their usage in large-scale musculoskeletal simulations. In this paper, we examine models of crossbridge dynamics of varying complexity and show that the emergent rheology of an ensemble of crossbridges can be simplified to a few dominant time-constants associated with intrinsic dynamical processes. For Huxley's classical two-state crossbridge model, we derive exact analytical expressions for the emergent ensemble rheology and find that it is characterized by a single time-constant. For more complex models with up to five crossbridge states, we show that at most three time-constants are needed to capture the ensemble rheology. Our results thus yield simplified models comprising of a few time-constants for muscle's bulk rheological response that can be readily used in large-scale simulations without sacrificing the model's interpretability in terms of the underlying actomyosin crossbridge dynamics.


2021 ◽  
Vol 17 (12) ◽  
pp. e1009761
Author(s):  
Yuzhen Liang ◽  
Chunwu Yu ◽  
Wentao Ma

The origin of life involved complicated evolutionary processes. Computer modeling is a promising way to reveal relevant mechanisms. However, due to the limitation of our knowledge on prebiotic chemistry, it is usually difficult to justify parameter-setting for the modeling. Thus, typically, the studies were conducted in a reverse way: the parameter-space was explored to find those parameter values “supporting” a hypothetical scene (that is, leaving the parameter-justification a later job when sufficient knowledge is available). Exploring the parameter-space manually is an arduous job (especially when the modeling becomes complicated) and additionally, difficult to characterize as regular “Methods” in a paper. Here we show that a machine-learning-like approach may be adopted, automatically optimizing the parameters. With this efficient parameter-exploring approach, the evolutionary modeling on the origin of life would become much more powerful. In particular, based on this, it is expected that more near-reality (complex) models could be introduced, and thereby theoretical research would be more tightly associated with experimental investigation in this field–hopefully leading to significant steps forward in respect to our understanding on the origin of life.


Sign in / Sign up

Export Citation Format

Share Document