• International journal of advanced smart convergence
• /
• 제5권3호
• /
• pp.40-46
• /
• 2016

The internet allows the information to flow at anywhere in anytime easily. Unfortunately, the network also becomes a great tool for the criminals to operate cybercrimes such as identity theft. To prevent the issue, using a very complex password is not a very encouraging method. Alternatively, keystroke dynamics helps the user to solve the problem. Keystroke dynamics is the information of timing details when a user presses a key or releases a key. A machine can learn a user typing behavior from the information integrate with a proper machine learning algorithm. In this paper, we have proposed mini-batch ensemble (MIBE) method which does the preprocessing on the original dataset and then produces multiple mini batches in the end. The mini batches are then trained by a machine learning algorithm. From the experimental result, we have shown the improvement of the performance for each base algorithm.

• 말소리와 음성과학
• /
• 제8권2호
• /
• pp.23-30
• /
• 2016

This paper describes some implementation schemes of CNN in view of mini-batch DNN training for efficient second order optimization. This uses same procedure updating parameters of DNN to train parameters of CNN by simply arranging an input image as a sequence of local patches, which is actually equivalent with mini-batch DNN training. Through this conversion, second order optimization providing higher performance can be simply conducted to train the parameters of CNN. In both results of image recognition on MNIST DB and syllable automatic speech recognition, our proposed scheme for CNN implementation shows better performance than one based on DNN.

• Industrial Engineering and Management Systems
• /
• 제10권1호
• /
• pp.1-6
• /
• 2011

This paper describes the issue of batch scheduling.In food production, the lead-time from produc-tion to sale should be decreased becausefreshness of the product is important. Products are shipped at diverse times depending on a demand of sellers, because the types of sellers has become diversified such as super-markets, convenience stores and etc. production of quantity demanded must be completed by time to ship it then. The authors consider a problem with due-dates constraints and construct the algorithm to find the opti-mal schedule that satisfy the due-dates constraint, batch size constraint, inventory time constraint and mini-mize total flow time.

• Journal of the Korean Data and Information Science Society
• /
• 제28권6호
• /
• pp.1245-1255
• /
• 2017

온라인 학습은 자료가 실시간으로 혹은 배치 단위로 축적되는 상황에서 주어진 목적함수의 해를 계산하는 방법을 말한다. 온라인 학습 알고리즘 중 배치를 이용한 확률적 경사 하강법 (stochastic gradient decent method)은 가장 많이 사용되는 방법 중 하나다. 이 방법은 구현이 쉬울 뿐만 아니라 자료가 동질적인 분포를 따른다는 가정 하에서 그 해의 성질이 잘 연구되어 있다. 하지만 자료에 특이값이 있거나 임의의 배치가 확률적으로 이질적 성질을 가질 때, 확률적 경사 하강법이 주는 해는 큰 편이를 가질 수 있다. 본 연구에서는 이러한 비정상 배치 (abnormal batch) 있는 자료 하에서 효과적으로 온라인 학습을 수행할 수 있는 수정된 경사 하강 알고리즘 (modified gradient decent algorithm)을 제안하고, 그 알고리즘을 통해 계산된 해의 수렴성을 밝혔다. 뿐만 아니라 간단한 모의실험을 통해 제안한 방법의 이론적 성질을 실증하였다.

• 스마트미디어저널
• /
• 제9권1호
• /
• pp.45-50
• /
• 2020

딥 러닝의 발전에 따라 추천시스템에서 딥 러닝 기반의 인공신경망을 활용한 연구가 활발히 진행되고 있다. 특히, RNN(Recurrent Neural Network)기반의 추천시스템은 데이터의 순차적 특성을 고려하기 때문에 추천시스템에서 좋은 성과를 보여주고 있다. 본 연구는 RNN기반의 알고리즘인 GRU(Gated Recurrent Unit)와 세션 기반 병렬 미니배치(Session Parallel mini-batch)기법을 활용한 여행경로 추천 시스템을 제안한다. 본 연구는 top1과 bpr(Bayesian personalized ranking) 오차함수의 앙상블을 통해 추천 성과를 향상시켰다. 또한, 데이터 내에 순차적인 특성을 고려한 RNN기반 추천 시스템은 여행경로에 내재된 여행지의 의미가 반영된 추천이 이루어진다는 것을 확인되었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권7호
• /
• pp.3714-3732
• /
• 2019

With the rapid development of network, Intrusion Detection System(IDS) plays a more and more important role in network applications. Many data mining algorithms are used to build IDS. However, due to the advent of big data era, massive data are generated. When dealing with large-scale data sets, most data mining algorithms suffer from a high computational burden which makes IDS much less efficient. To build an efficient IDS over big data, we propose a classification algorithm based on data clustering and data reduction. In the training stage, the training data are divided into clusters with similar size by Mini Batch K-Means algorithm, meanwhile, the center of each cluster is used as its index. Then, we select representative instances for each cluster to perform the task of data reduction and use the clusters that consist of representative instances to build a K-Nearest Neighbor(KNN) detection model. In the detection stage, we sort clusters according to the distances between the test sample and cluster indexes, and obtain k nearest clusters where we find k nearest neighbors. Experimental results show that searching neighbors by cluster indexes reduces the computational complexity significantly, and classification with reduced data of representative instances not only improves the efficiency, but also maintains high accuracy.

• 한국멀티미디어학회논문지
• /
• 제23권8호
• /
• pp.977-985
• /
• 2020

Aiming at the problem of complex feature extraction and low accuracy in human action recognition, this paper proposed a network structure combining batch normalization algorithm with GoogLeNet network model. Applying Batch Normalization idea in the field of image classification to action recognition field, it improved the algorithm by normalizing the network input training sample by mini-batch. For convolutional network, RGB image was the spatial input, and stacked optical flows was the temporal input. Then, it fused the spatio-temporal networks to get the final action recognition result. It trained and evaluated the architecture on the standard video actions benchmarks of UCF101 and HMDB51, which achieved the accuracy of 93.42% and 67.82%. The results show that the improved convolutional neural network has a significant improvement in improving the recognition rate and has obvious advantages in action recognition.

• 한국산학기술학회논문지
• /
• 제12권9호
• /
• pp.4261-4266
• /
• 2011

폴리프로필렌(polypropylene, PP)/talc 컴파운드를 talc의 조성을 20wt%로 하여 마스터뱃취(master-batch, MB)와 사출기를 이용한 직접 제조법으로 각각 제조하였다. MB는 talc 함량을 50wt%로 하여 소형 혼합기(mini compounder)로 $200^{\circ}C$에서 제조하였다. SEM-EDS를 이용하여 PP/talc 컴파운드의 talc 분산정도를 평가하였다. MB를 이용하여 제조한 컴파운드의 talc 분산성이 우수하게 나타났다. PP/talc 컴파운드의 유변학적 특성을 동적유변측정기를 이용하여 측정하였다. MB를 이용한 MB-PP 컴파운드가 직접 제조한 컴파운드보다 shear thinning과 탄성특성이 증가하였다. G'-G" plot의 기울기 변화로부터 talc 분산성 정도를 확인하였고, 탄성특성 증가를 확인하기 위해 Van Gurp-Palmen 분석을 이용하였다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2023년도 춘계학술발표대회
• /
• pp.585-587
• /
• 2023

Long-tailed datasets have an imbalanced distribution because they consist of a different number of data samples for each class. However, there are problems of the performance degradation in tail-classes and class-accuracy imbalance for all classes. To address these problems, this paper suggests a learning method for training of long-tailed dataset. The proposed method uses and combines two methods; one is a resampling method to generate a uniform mini-batch to prevent the performance degradation in tail-classes, and the other is a reweighting method to address the accuracy imbalance problem. The purpose of our proposed method is to train the learning models to have uniform accuracy for each class in a long-tailed dataset.

• 한국과학교육학회지
• /
• 제28권7호
• /
• pp.730-748
• /
• 2008

Students often have tenacious physics misconceptions and many studies were conducted on engendering conceptual change. Correspondingly, there is much literature on alternative assessment and its role in student learning. This is a comparison study on using alternative assessments to improve common students' misconceptions in GCE Advanced Level Physics. This research also aims to affirm alternative assessment as a valid tool for learning and promote its use. This study involved two classes with 24 students each. For four weeks, electromagnetism was taught to students using the same classroom pedagogies but with different assignments. The control group completeda standard drill-and-practice assignment while the experimental group finished an alternative assessment. From the preliminary results, students who undertook the alternative assessment and the traditional assessment both improved, however, the treatment group did not perform statistically significantly better than the control group. The reasons will be discussed and commented and it is expected to have significant improvement on rectifying misconceptionsupon next batch of experimentation groups.