• Journal of Platform Technology
• /
• 제11권1호
• /
• pp.72-84
• /
• 2023

With the rise of the Internet of Things, the security of such lightweight computing environments has become a hot topic. Lightweight block ciphers that can provide efficient performance and security by having a relatively simpler structure and smaller key and block sizes are drawing attention. Due to these characteristics, they can become a target for new attack techniques. One of the new cryptanalytic attacks that have been attracting interest is Neural cryptanalysis, which is a cryptanalytic technique based on neural networks. It showed interesting results with better results than the conventional cryptanalysis method without a great amount of time and cryptographic knowledge. The first work that showed good results was carried out by Aron Gohr in CRYPTO'19, the attack was conducted on the lightweight block cipher SPECK-/32/64 and showed better results than conventional differential cryptanalysis. In this paper, we first apply the Differential Neural Distinguisher proposed by Aron Gohr to the block ciphers HIGHT and GOST to test the applicability of the attack to ciphers with different structures. The performance of the Differential Neural Distinguisher is then analyzed by replacing the neural network attack model with five different models (Multi-Layer Perceptron, AlexNet, ResNext, SE-ResNet, SE-ResNext). We then propose a Related-key Neural Distinguisher and apply it to the SPECK-/32/64, HIGHT, and GOST block ciphers. The proposed Related-key Neural Distinguisher was constructed using the relationship between keys, and this made it possible to distinguish more rounds than the differential distinguisher.

• 한국수자원학회논문집
• /
• 제55권11호
• /
• pp.903-911
• /
• 2022

물을 공급하기 위한 자원 중 하나인 지하수는 다양한 자연적 요인에 의해 수위의 변동이 발생한다. 최근, 인공신경망을 이용하여 지하수위의 변동을 예측하는 연구가 진행되었다. 기존에는 인공신경망 연산자 중 학습에 영향을 미치는 Optimizer로 경사하강법(Gradient Descent, GD) 기반 Optimizer를 사용하였다. GD 기반 Optimizer는 초기 상관관계 의존성과 해의 비교 및 저장 구조 부재의 단점이 존재한다. 본 연구는 GD 기반 Optimizer의 단점을 개선하기 위해 GD와 화음탐색법(Harmony Search, HS)를 결합한 새로운 Optimizer인 Gradient Descent combined with Harmony Search(GDHS)를 개발하였다. GDHS의 성능을 평가하기 위해 다층퍼셉트론(Multi Layer Perceptron, MLP)을 이용하여 이천율현 관측소의 지하수위를 학습 및 예측하였다. GD 및 GDHS를 사용한 MLP의 성능을 비교하기 위해 Mean Squared Error(MSE) 및 Mean Absolute Error(MAE)를 사용하였다. 학습결과를 비교하면, GDHS는 GD보다 MSE의 최대값, 최소값, 평균값 및 표준편차가 작았다. 예측결과를 비교하면, GDHS는 GD보다 모든 평가지표에서 오차가 작은 것으로 평가되었다.

• 드라이브 ㆍ 컨트롤
• /
• 제20권4호
• /
• pp.133-139
• /
• 2023

A weighing system calculates the bucket's excavation amount of an excavator. Usually, the excavation amount is computed by the excavator's motion equations with sensing data. But these motion equations have computing errors that are induced by assumptions to the linear systems and identification of the equation's parameters. To reduce computing errors, some commercial weighing system incorporates particular motion into the excavation process. This study introduces a linear regression model on an artificial neural network that has fewer predicted errors and doesn't need a particular pose during an excavation. Time serial data were gathered from a 30tons excavator's loading test. Then these data were preprocessed to be adjusted by MPL (Multi Layer Perceptron) or CNN (Convolutional Neural Network) based linear regression models. Each model was trained by changing hyperparameter such as layer or node numbers, drop-out rate, and kernel size. Finally ID-CNN-based linear regression model was selected.

• Smart Structures and Systems
• /
• 제30권2호
• /
• pp.195-207
• /
• 2022

Computational drawbacks associated with regular predictive models have motivated engineers to use hybrid techniques in dealing with complex engineering tasks like simulating the compressive strength of concrete (CSC). This study evaluates the efficiency of tree potential metaheuristic schemes, namely shuffled complex evolution (SCE), multi-verse optimizer (MVO), and beetle antennae search (BAS) for optimizing the performance of a multi-layer perceptron (MLP) system. The models are fed by the information of 1030 concrete specimens (where the amount of cement, blast furnace slag (BFS), fly ash (FA1), water, superplasticizer (SP), coarse aggregate (CA), and fine aggregate (FA2) are taken as independent factors). The results of the ensembles are compared to unreinforced MLP to examine improvements resulted from the incorporation of the SCE, MVO, and BAS. It was shown that these algorithms can considerably enhance the training and prediction accuracy of the MLP. Overall, the proposed models are capable of presenting an early, inexpensive, and reliable prediction of the CSC. Due to the higher accuracy of the BAS-based model, a predictive formula is extracted from this algorithm.

• 전자공학회논문지B
• /
• 제32B권5호
• /
• pp.765-776
• /
• 1995

This paper describes a speaker adaptation method to improve the recognition performance of MLP(multiLayer Perceptron) based HMM(Hidden Markov Model) speech recognizer. In this method, we use lst-order linear transformation network to fit data of a new speaker to the MLP. Transformation parameters are adjusted by back-propagating classification error to the transformation network while leaving the MLP classifier fixed. The recognition system is based on semicontinuous HMM's which use the MLP as a fuzzy vector quantizer. The experimental results show that rapid speaker adaptation resulting in high recognition performance can be accomplished by this method. Namely, for supervised adaptation, the error rate is signifecantly reduced from 9.2% for the baseline system to 5.6% after speaker adaptation. And for unsupervised adaptation, the error rate is reduced to 5.1%, without any information from new speakers.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2000년도 ITC-CSCC -2
• /
• pp.1041-1044
• /
• 2000

The enhancement to the back-propagation algorithm presented in this paper has resulted from the need to extract sparsely connected networks from networks employing product terms. The enhancement works in conjunction with the back-propagation weight update process, so that the actions of weight zeroing and weight stimulation enhance each other. It is shown that the error measure, can also be interpreted as rate of weight change (as opposed to ${\Delta}W_{ij}$), and consequently used to determine when weights have reached a stable state. Weights judged to be stable are then compared to a zero weight threshold. Should they fall below this threshold, then the weight in question is zeroed. Simulation of such a system is shown to return improved learning rates and reduce network connection requirements, with respect to the optimal network solution, trained using the normal back-propagation algorithm for Multi-Layer Perceptron (MLP), Higher Order Neural Network (HONN) and Sigma-Pi networks.

• Journal of Electrical Engineering and Technology
• /
• 제5권4호
• /
• pp.666-671
• /
• 2010

Multi-layer perceptron neural networks have been constructed to classify four types of defects in steam generator tubes. Three features are extracted from the signals of the eddy current testing method. These include maximum impedance, phase angle at the point of maximum impedance, and an angle between the point of maximum impedance and the point of half the maximum impedance. Two hundred sets of these features are used for training and assessing the networks. Two approaches are involved to train the networks and to classify the defect type. One is the conjugate gradient method and the other is the Broydon-Fletcher-Goldfarb-Shanno method which is recognized as the most popular algorithm of quasi-Newton methods. It is found from the computation results that the training time of the Broydon-Fletcher-Goldfarb-Shanno method is much faster than that of the conjugate gradient method in most cases. On the other hand, no significant difference of the classification performance between the two methods is observed.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2005년도 추계학술대회 학술발표 논문집 제15권 제2호
• /
• pp.483-487
• /
• 2005

We introduce a new architecture of hetero-hybridized feed-forward neural networks composed of fuzzy set-based polynomial neural networks (FSPNN) and polynomial neural networks (PM) that are based on a genetically optimized multi-layer perceptron and develop their comprehensive design methodology involving mechanisms of genetic optimization and Information Granulation. The construction of Information Granulation based HFSPNN (IG-HFSPNN) exploits fundamental technologies of Computational Intelligence(Cl), namely fuzzy sets, neural networks, and genetic algorithms(GAs) and Information Granulation. The architecture of the resulting genetically optimized Information Granulation based HFSPNN (namely IG-gHFSPNN) results from a synergistic usage of the hybrid system generated by combining new fuzzy set based polynomial neurons (FPNs)-based Fuzzy Neural Networks(PM) with polynomial neurons (PNs)-based Polynomial Neural Networks(PM). The design of the conventional genetically optimized HFPNN exploits the extended Group Method of Data Handling(GMDH) with some essential parameters of the network being tuned by using Genetie Algorithms throughout the overall development process. However, the new proposed IG-HFSPNN adopts a new method called as Information Granulation to deal with Information Granules which are included in the real system, and a new type of fuzzy polynomial neuron called as fuzzy set based polynomial neuron. The performance of the IG-gHFPNN is quantified through experimentation.

• 정보보호학회논문지
• /
• 제33권2호
• /
• pp.193-200
• /
• 2023

암호 분석은 알려진 평문 공격, 차분 분석, 부채널 분석 등과 같이 다양한 기법으로 수행될 수있다. 최근에는 딥러닝을 암호 분석에 적용하는 연구들이 제안되고 있다. 알려진 평문 공격(Known-plaintext Attack)은 알려진 평문과 암호문 쌍을 사용하여 키를 알아내는 암호 분석 기법이다. 본 논문에서는 딥러닝 기술을 사용하여 경량 블록 암호 PRESENT의 축소 버전인 S-PRESENT에 대해 알려진 평문 공격을 수행한다. 축소된 경량 블록 암호에 대해 수행된 최초의 딥러닝 기반의 알려진 평문 공격이라는 점에서 본 논문은 의의가 있다. 성능 향상 및 학습속도 개선을 위해 Skip connection, 1x1 Convolution과 같은 딥러닝 기법을 적용하였다. 암호 분석에는 MLP(Multi-Layer Perceptron)와 1D, 2D 합성곱 신경망 모델을 사용하여 최적화하였으며, 세 모델의 성능을 비교한다. 2D 합성곱 신경망에서 가장 높은 성능을 보였지만 일부 키공간까지만 공격이 가능했다. 이를 통해 MLP 모델과 합성곱 신경망을 통한 알려진 평문 공격은 공격 가능한 키 비트에 제한이 있음을 알 수 있다.

• 한국통신학회논문지
• /
• 제24권12B호
• /
• pp.2447-2454
• /
• 1999

본 논문은 기존 분류형 신경망의 인식성능을 향상시키기 위하여 프레임 정규화와 비선형 사후확률 추정법(N-APPEM)을 제안하고 한국어 숫자음에 대하여 예측형과 분류형 신경망으로 인식성능을 평가하였다. 실험결과 예측형 신경망에서 최고 98.0%의 인식률을 얻었다. 예측형 신경망은 네트워크가 입력패턴의 카테고리 수만큼 마련되는 복잡한 네트워크를 가지는 반면에 분류형 신경망은 단일 네트워크로 구성되며 프레임 정규화와 비선형 사후확률 추정법으로 85.5%까지 인식률을 향상시킬 수 있었으며 이는 기존의 방법보다 인식률이 12.0% 향상된 것이다.