• International Journal of Computer Science & Network Security
• /
• v.23 no.8
• /
• pp.177-189
• /
• 2023

Malware detection is an increasingly important operational focus in cyber security, particularly given the fast pace of such threats (e.g., new malware variants introduced every day). There has been great interest in exploring the use of machine learning techniques in automating and enhancing the effectiveness of malware detection and analysis. In this paper, we present a deep recurrent neural network solution as a stacked Long Short-Term Memory (LSTM) with a pre-training as a regularization method to avoid random network initialization. In our proposal, we use global and short dependencies of the inputs. With pre-training, we avoid random initialization and are able to improve the accuracy and robustness of malware threat hunting. The proposed method speeds up the convergence (in comparison to stacked LSTM) by reducing the length of malware OpCode or bytecode sequences. Hence, the complexity of our final method is reduced. This leads to better accuracy, higher Mattews Correlation Coefficients (MCC), and Area Under the Curve (AUC) in comparison to a standard LSTM with similar detection time. Our proposed method can be applied in real-time malware threat hunting, particularly for safety critical systems such as eHealth or Internet of Military of Things where poor convergence of the model could lead to catastrophic consequences. We evaluate the effectiveness of our proposed method on Windows, Ransomware, Internet of Things (IoT), and Android malware datasets using both static and dynamic analysis. For the IoT malware detection, we also present a comparative summary of the performance on an IoT-specific dataset of our proposed method and the standard stacked LSTM method. More specifically, of our proposed method achieves an accuracy of 99.1% in detecting IoT malware samples, with AUC of 0.985, and MCC of 0.95; thus, outperforming standard LSTM based methods in these key metrics.

• Journal of Advanced Marine Engineering and Technology
• /
• v.21 no.2
• /
• pp.195-205
• /
• 1997

The propulsion marine diesel engine have been widely applied with a mechanical- hydraulic governor to control the ship speed for long time. But it was recently very difficult for the mechani¬cal - hydraulic governor to control the speed of engine under the condition of low speed and low load because of jiggling by rough fluctuation of rotating torque and hunting by dead time of Desiel engnie The performance improvement of mechanical - hydraulic governor is required to solve these problems of control system. The electro - hydraulic governor using PID algorithm is provided to compensate the faults of mechanical- hydraulic governor. In this paper, in order to analyze the ship speed control system, the transfer function was converted from the z tansformation to w transformation. The influence of dead time changing by engine speed which induces hunting phenomena was investigated by Nichols chart of w plane. As a method of performance improvement of mechanical hydraulic governor, a Eletro - hydraulic governor shows that fine control results can be obtained through optimal parameter tuning of PID

• Advances in Computational Design
• /
• v.1 no.2
• /
• pp.175-194
• /
• 2016

Nature has provided inspiration for most of the man-made technologies. Scientists believe that dolphins are the second to humans in smartness and intelligence. Echolocation is the biological sonar used by dolphins for navigation and hunting in various environments. This ability of dolphins is mimicked in this paper to develop a new optimization method. Dolphin Echolocation Optimization (DEO) is an optimization method based on dolphin's approach for hunting food and exploration of environment. DEO has already been developed for discrete optimization search space and here it is extended to continuous search space. DEO has simple rules and is adjustable for predetermined computational cost. DEO provides the optimum results and leads to alternative optimality curves suitable for the problem. This algorithm has a few parameters and it is applicable to a wide range of problems like other metaheuristic algorithms. In the present work, the efficiency of this approach is demonstrated using standard benchmark problems.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.496-498
• /
• 1998

This paper presents a mathod of reducing hunting size or steady state error occurred in the output signals via regulating the PID controllers gains. The PID controllers are widely used in industrial processes. Such processes have several inherent features like continuous operation, fixed set value, and difficulty in applyirty test signals. Thus, this paper suggests fuzzy rules of reducing hunting magnitude or steady state error using output signals only. Such an intelligent tuning technique utilizes both the experts, experience and control engineers' theortical background. For two kinds of systems such as temperature or DC motors speed control, we showed the validity of proposed method in this paper.

• The Research Journal of the Costume Culture
• /
• v.13 no.5 s.58
• /
• pp.699-711
• /
• 2005

This study was following one of 'A study on the consciousness of fashion industries internship'. The purposes of this study were to develop the internship program which focused on practical use to introduce and carry out for fashion industries, and secondly to propose some regime for government to activate fashion internship. Reference searching method and depth interviewing method were used for this study. The results were as follows : Fashion industry internship was grouped into two classes, 'on-the-job training'; educating students fields and 'talent hunting'; selecting good persons. Internship of industry-academic world was classified into two types; the one is 'credit type' which has curriculums between universities and industries and the other is 'non-credit type' which has not any credit and is operated by industry own system. This study provided the development courses of pragmatic program to perform internship systematically and it also provided the program models for guide line in fashion industries. Six grades such as ready step, introduction step, selection step, management step, evaluation step and feed-back step were proposed for the internship program development steps of fashion industries. A virtual organization, 'The Fashion Industry and Academy Association' was proposed as a policy for activating internship between universities, industries and government.

• 제어로봇시스템학회:학술대회논문집
• /
• 1992.10a
• /
• pp.937-940
• /
• 1992

A cost-saving feeding rate contro method for a batch reaction process is proposed by reconfigurating the flow rate control loop where the feeding rate is estimated by differentiating weight measurement. While the existing control methods require either a flow rate measurement device or a flow rate estimator, the proposed method is devised not to require neither of these facilities by directly controlling the tank weight with a time-varying set point. Experimental evaluation in a pilot-scale unit shows that the proposed method can significantly reduce the actuator hunting of the existing method where the flow rate is estimated by differentiating the weight measurement.

• Proceedings of the Materials Research Society of Korea Conference
• /
• 2005.05a
• /
• pp.200-200
• /
• 2005

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.10 no.5
• /
• pp.742-750
• /
• 1986

The lateral vibrating motion of a railway vehicle over a certain critical speed is a well known problem in the field of train dynamics. It is known that the train equations of motion are strongly coupled and highly nonlinear with the motion and causing that it is very difficult to solve the equations simultaneously. In this paper, a 8 degree of feedom model of a railway vehicle was suggested to solve the rail vehicle lateral motion. In stead of solving the nonlinear equation simultaneously, statistical linearization technique was adopted to solve those equations. The analysis results from the statistical linearization method were directly compared with those from direct nonlinear equations and found that the linearization technique can be very effective and economical for railroad vehicle analysis. By the way, it was found that the analysis results can analytically explain the intermittent hunting phenomena which has been frequently observed in experiments.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.68 no.2
• /
• pp.262-268
• /
• 2019

Previous dynamic models of hysteresis motor use an extended induction machine equivalent circuit or somewhat different equivalent circuit with conventional one, which makes unsatisfiable results. In this paper, the hysteresis dynamic characteristics of the motor rotor are analyzed using the inverse Preisach model and the hysteresis motor equivalent circuit considering eddy current effect. The hysteresis loop for the rotor ring is analyzed under full-load voltage source static state. The calculated hysteresis loop is then approximated to an ellipse for simplicity of dynamic computation. The permeability and delay angle of the elliptic loop apply to the dynamic analysis model. As a result, it is possible to dynamically analyze the hysteresis motor according to the applied voltage and the rotor material. With this method, the motor speed, generated torque, load angle, rotor current as well as synchronous entry time, hunting effect can be calculated.

• Journal of Power Electronics
• /
• v.18 no.3
• /
• pp.841-852
• /
• 2018

This paper presents a dual-algorithm search method (GWO-GSO) combining grey wolf optimization (GWO) and golden-section optimization (GSO) to realize maximum power point tracking (MPPT) for photovoltaic (PV) systems. First, a modified grey wolf optimization (MGWO) is activated for the global search. In conventional GWO, wolf leaders possess the same impact on decision-making. In this paper, the decision weights of wolf leaders are automatically adjusted with hunting progression, which is conducive to accelerating hunting. At the later stage, the algorithm is switched to GSO for the local search, which play a critical role in avoiding unnecessary search and reducing the tracking time. Additionally, a novel restart judgment based on the quasi-slope of the power-voltage curve is introduced to enhance the reliability of MPPT systems. Simulation and experiment results demonstrate that the proposed algorithm can track the global maximum power point (MPP) swiftly and reliably with higher accuracy under various conditions.