• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.2
• /
• pp.456-477
• /
• 2024

With information technology's rapid development, the Internet faces serious security problems. Studies have shown that malware has become a primary means of attacking the Internet. Therefore, adversarial samples have become a vital breakthrough point for studying malware. By studying adversarial samples, we can gain insights into the behavior and characteristics of malware, evaluate the performance of existing detectors in the face of deceptive samples, and help to discover vulnerabilities and improve detection methods for better performance. However, existing adversarial sample generation methods still need help regarding escape effectiveness and mobility. For instance, researchers have attempted to incorporate perturbation methods like Fast Gradient Sign Method (FGSM), Projected Gradient Descent (PGD), and others into adversarial samples to obfuscate detectors. However, these methods are only effective in specific environments and yield limited evasion effectiveness. To solve the above problems, this paper proposes a malware adversarial sample generation method (PixGAN) based on the pixel attention mechanism, which aims to improve adversarial samples' escape effect and mobility. The method transforms malware into grey-scale images and introduces the pixel attention mechanism in the Deep Convolution Generative Adversarial Networks (DCGAN) model to weigh the critical pixels in the grey-scale map, which improves the modeling ability of the generator and discriminator, thus enhancing the escape effect and mobility of the adversarial samples. The escape rate (ASR) is used as an evaluation index of the quality of the adversarial samples. The experimental results show that the adversarial samples generated by PixGAN achieve escape rates of 97%, 94%, 35%, 39%, and 43% on the Random Forest (RF), Support Vector Machine (SVM), Convolutional Neural Network (CNN), Convolutional Neural Network and Recurrent Neural Network (CNN_RNN), and Convolutional Neural Network and Long Short Term Memory (CNN_LSTM) algorithmic detectors, respectively.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.1273-1276
• /
• 1998

The HMM-Net is an architecture for a neural network that implements a hidden Markov model (HMM). The architecture is developed for the purpose of combining the discriminant power of neural networks with the time-domain modeling capability of HMMs. Criteria used for learning HMM-Net classifiers are maximum likelihood (ML), maximum mutual information (MMI), and minimization of mean squared error(MMSE). In this paper we propose an efficient learning method of HMM-Net classifiers using hybrid criteria, ML/MMSE and MMI/MMSE, and report the results of an experimental study comparing the performance of HMM-Net classifiers trained by the gradient descent algorithm with the above criteria. Experimental results for the isolated numeric digits from /0/ to /9/ show that the performance of the proposed method is better than the others in the respects of learning and recognition rates.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.24 no.10
• /
• pp.1335-1342
• /
• 2000

Three-dimensional flow analysis and numerical optimization methods are presented for the design of an axial-flow fan. Steady, incompressible, three-dimensional Reynolds-averaged Navier-Stokes equations are used as governing equations, and standard k- ${\varepsilon}$ turbulence model is chosen as a turbulence model. Governing equations are discretized using finite volume method. Steepest descent method, conjugate gradient method and BFGS method are compared to determine the searching directions. Golden section method and quadratic fit-sectioning method are tested for one dimensional search. Objective function is defined as a ratio of generation rate of the turbulent kinetic energy to pressure head. Two variables concerning sweep angle distribution are selected as the design variables. Performance of the final fan designed by the optimization was tested experimentally.

• Proceedings of the KIEE Conference
• /
• 2007.11c
• /
• pp.70-73
• /
• 2007

Many methods have been developed to overcome the PAPR(peak-to-average power ratio) problem. Selective mapping(SLM) partial transmit sequence(PTS), subblock phase weighting(SPW) and gradient descent(GD) are used widely to reduce the PAPR. In this paper, we present a effective PAPR reduction method, decrease the calculation through Radix-2 DIF IFFT procedure and GD method. and can transmit selecting data sequence that satisfy threshold value as one part of proposed method, in case satisfy fixed threshold value, or transmit selecting data sequence with the lower papr operating remained part for performance improvement.

• Proceedings of the KIEE Conference
• /
• 2000.11d
• /
• pp.803-805
• /
• 2000

In this paper, we propose fuzzy-neural Networks(FNN) which is useful for identification algorithms. The proposed FNN model consists of two steps: the first step, which determines premise and consequent parameters approximately using FCM_RI method, the second step, which adjusts the premise and consequent parameters more precisely by gradient descent algorithm. The FCM_RI algorithm consists FCM clustering algorithm and Recursive least squared(RLS) method, this divides the input space more efficiently than convention methods by taking into consideration correlations between components of sample data. To evaluate the performance of the proposed FNN model, we use the time series data for gas furnace.

• Journal of Korea Multimedia Society
• /
• v.17 no.12
• /
• pp.1437-1445
• /
• 2014

A smartphone has very limited input methods regardless of its various functions. In this respect, it is one alternative that sensor motion recognition can make intuitive and various user interface. In this paper, we recognize user's motion using acceleration sensor, magnetic field sensor, and gyro sensor in smartphone. We try to reduce sensing error by gradient descent algorithm because in single sensor it is hard to obtain correct data. And we apply vector quantization by conversion of rotation displacement to spherical coordinate system for elevated recognition rate and recognition of small motion. After vector quantization process, we recognize motion using HMM(Hidden Markov Model).

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.7 no.9
• /
• pp.2284-2298
• /
• 2013

In this paper, a fast and efficient signal reconstruction algorithm for solving the basis pursuit (BP) problem in compressed sensing (CS) is proposed. This fast linearized Bregman method (FLBM), which is inspired by the fast method of Beck et al., is based on the fact that the linearized Bregman method (LBM) is equivalent to a gradient descent method when applied to a certain formulation. The LBM requires $O(1/{\varepsilon})$ iterations to obtain an ${\varepsilon}$-optimal solution while the FLBM reduces this iteration complexity to $O(1/\sqrt{\varepsilon})$ and requiring almost the same computational effort on each iteration. Our experimental results show that the FLBM can be faster than some other existing signal reconstruction methods.

• Journal of Power Electronics
• /
• v.16 no.6
• /
• pp.2150-2161
• /
• 2016

Mechanical and electrical parameter uncertainties cause dynamic and static estimation errors of the rotor speed and position, resulting in performance deterioration of sensorless control systems. This paper applies an extended nonlinear observer to interior permanent magnet synchronous motors (IPMSM) for the simultaneous estimation of the rotor speed and position. Two compensation methods are proposed to improve the observer performance against parameter uncertainties: an on-line rotational inertia adjustment approach that employs the gradient descent algorithm to suppress dynamic estimation errors, and an equivalent flux error compensation approach to eliminate static estimation errors caused by inaccurate electrical parameters. The effectiveness of the proposed control strategy is demonstrated by experimental tests.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.8
• /
• pp.211-218
• /
• 1994

In this paper we develope a hybrid learning rule to improve the robustness of multi-layer Perceptions. In most neural networks the activation of a neuron is deternined by a nonlinear transformation of the weighted sum of inputs to the neurons. Investigating the behaviour of activations of hidden layer neurons a new learning algorithm is developed for improved robustness for multi-layer Perceptrons. Unlike other methods which reduce the network complexity by putting restrictions on synaptic weights our method based on error-backpropagation increases the complexity of the underlying proplem by imposing it saturation requirement on hidden layer neurons. We also found that the additional gradient-descent term for the requirement corresponds to the Hebbian rule and our algorithm incorporates the Hebbian learning rule into the error back-propagation rule. Computer simulation demonstrates fast learning convergence as well as improved robustness for classification and hetero-association of patterns.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.537-539
• /
• 1998

In this paper, we construct the on-line model structure for the nonlinear process systems using the adaptive fuzzy-neural network. Adaptive fuzzy-neural network usually consists of two distinct modifiable structure, with both, the premise and the consequent part. These two parts can be adapted by different optimization methods, which are the hybrid learning procedure combining gradient descent method and least square method. To achieve the on-line model structure, we use the recursive least square method for the consequent parameter identification of nonlinear process. We design the interface between PLC and main computer, and construct the monitoring and control simulator for the nonlinear process. The proposed on-line modeling to real process is carried out to obtain the effective and accurate results.