• The KIPS Transactions:PartD
• /
• v.13D no.2 s.105
• /
• pp.199-206
• /
• 2006

Sequential pattern mining is an important data mining task with broad applications. However, conventional methods may meet inherent difficulties in mining databases with long sequences and noise. They may generate a huge number of short and trivial patterns but fail to find interesting patterns shared by many sequences. In this paper, to overcome these problems, we propose the theme of approximate sequential pattern mining roughly defined as identifying patterns approximately shared by many sequences. The proposed method works in two steps: one is to cluster target sequences by their similarities and the other is to find consensus patterns that ire similar to the sequences in each cluster directly through multiple alignment. For this purpose, a novel structure called weighted sequence is presented to compress the alignment result, and the longest consensus pattern that represents each cluster is generated from its weighted sequence. Finally, the effectiveness of the proposed method is verified by a set of experiments.

• The KIPS Transactions:PartD
• /
• v.10D no.6
• /
• pp.1025-1032
• /
• 2003

Percentiles are convenient measures of the entire range of values of simulation outputs. However, unlike means and standard deviations, the observations have to be stored since calculation of percentiles requires several passes through the data. Thus, percentile (PE) requires a large amount of computer storage and computation time. The best possible computation time to sort n observations is (O($nlog_{2}n$)), and memory proportional to n is required to store sorted values in order to find a given order statistic. Several approaches for extimating percentiles in RS(regenerative simulation) and non-RS, which can avoid difficulties of PE, have been proposed in [11, 12, 21]. In this paper, we implemented these three approaches known as : leanear PE, batching PE, spectral $P^2$ PE in the context of sequential steady-state simulation. Numerical results of coverage analysis of these PE approachs are present.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 1990.10a
• /
• pp.1-16
• /
• 1990

In this paper, a system to measure treeing degradation phenomena in organic insulating materials, using an image sensor, is discussed. Using an image processing technique, tree features immediately after tree initiation as well as changes in the configuration of the tree were measured. which up to now have been extremely difficult to observe by conventional visual methods. As a result, it was possible to record the image of tree propagation immediately after its first appearance, and to describe the specific characteristics of tree growth in terms of the length, the degraded area and the sequential images.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.362-364
• /
• 2018

핸드폰 사용자가 늘어나면서 이와 관련하여 개인 정보 보안에 대한 중요성이 대두되고 있다. 이에 따라 제안된 알고리즘은 Extreme learning machine 으로부터 착안하여 변형하여 고안한 Maxtreme Learning Machine(MLM) 으로, 사용자들의 터치 스트로크 특성 벡터를 제안 알고리즘으로 학습하여 사용자들을 검증한다. 또한 특성 벡터의 순차적 융합 기법을 이용하여 더 많은 정보를 바탕으로 사용자를 높은 정확도로 검증 할 수 있다.

• ETRI Journal
• /
• v.5 no.4
• /
• pp.17-21
• /
• 1983

Sine call handling for stored program control exchange can be regarded as finite state machine model of sequential process, the state transition concept is applied to describe its functions. Coding method of the state transition table applied to M10CN ESS call processing and concerned topics was discussed and proposed another method for reduction memory usage.

• The KIPS Transactions:PartB
• /
• v.11B no.5
• /
• pp.603-610
• /
• 2004

The distributed Information Retrieval System centrally controlled by mediator or meta search engine result in congestion of heavy traffic and int he problem of increment of cost for the reason of the design of complicated algorithm for central control and installation of hardware. So to figure out this problem, the way is needed that has independent retrieval functionality and can cooperate each other without dependency. In this paper, we overview a few works involved in distributed information retrieval system, then, implement algorithm and design the frame-work of distributed information retrieval system using sequential transferring method(STM) including multiple information retrieval system separated from central control. For this first of all, we present a web partition policy which devide and manage web logically and we present the sequential query processing way by means of illustration through changing numbered information retrieval system. Then, we also present 3-layered structure of framework and function and module of each layer suitable for information retrieval system. Last of ail, for effective implementation of STM algorithm we analysis module structure and present description of pseudocode of this, and show that the proposed STM algorithm works smoothly by demonstration of sequential query transfer process between servers.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.9 no.5
• /
• pp.205-211
• /
• 2009

The Weightless Neural Network(WNN) has the advantage of the processing speed, less computability than weighted neural network which readjusts the weight. Especially, The behavior information such as sequential gesture has many serial correlation. So, It is required the high computability and processing time to recognize. To solve these problem, Many algorithms used that added preprocessing and hardware interface device to reduce the computability and speed. In this paper, we proposed the Ram based Sequential Cumulative Neural Network(SCNN) model which is sign language recognition system without preprocessing and hardware interface. We experimented with using compound words in continuous korean sign language which was input binary image with edge detection from camera. The recognition system of sign language without preprocessing got 93% recognition rate.

• The Transactions of the Korean Institute of Electrical Engineers C
• /
• v.52 no.6
• /
• pp.235-240
• /
• 2003

In this paper, we present the fabrication and the characteristic analysis of sequential lateral solidification(SLS) poly-Si thin film transistors(TFT's) with molybdenum gate for active matrix liquid displays (AMLCD's) pixel controlling devices. The molybdenum gate is applied for the purpose of low temperature processing. The maximum processing temperature is 55$0^{\circ}C$ at the dopant thermal annealing step. The SLS processed poly-Si film which is reduced grain and grain boundary effect, is applied for the purpose of electrical characteristics improvements of poly-Si TFT's. The fabricated low temperature SLS poly-Si TFT's had a varying the channel length and width from 10${\mu}{\textrm}{m}$ to 2${\mu}{\textrm}{m}$. And to analyze these devices, extract electrical characteristic parameters (field effect mobility, threshold voltage, subthreshold slope, on off current etc) from current-voltage transfer characteristics curve. The extract electrical characteristic of fabricated low temperature SLS poly-Si TFT's showed the mobility of 100~400cm$^2$/Vs, the off current of about 100pA, and the on/off current ratio of about $10^7$. Also, we observed that the change of grain boundary according to varying channel length is dominant for the change of electrical characteristics more than the change of grain boundary according to varying channel width. Hereby, we comprehend well the characteristics of SLS processed poly-Si TFT's witch is recrystallized to channel length direction.

• The Transactions of the Korea Information Processing Society
• /
• v.5 no.6
• /
• pp.1620-1632
• /
• 1998

This paper presents a fast parallel algorithm for volume rendering and its implementation using C language and MPI MasPar Programming Language) on the 4,096 processor MasPar MP-2 machine. This parallel algorithm is a parallelization hased on the Lacroute' s sequential shear - warp algorithm currently acknowledged to be the fastest sequential volume rendering algorithm. This algorithm reduces communication overheads by using the sheared space partition scheme and the load balancing technique using load estimates from the previous iteration, and the number of voxels to be processed by using the run-length encoded volume data structure.Actual performance is 3 to 4 frames/second on the human hrain scan dataset of $128\times128\times128$ voxels. Because of the scalability of this algorithm, performance of ]2-16 frames/sc.'cond is expected on the 16,384 processor MasPar MP-2 machine. It is expected that implementation on more current SIMD or MIMD architectures would provide 3O~60 frames/second on large volumes.

• The KIPS Transactions:PartD
• /
• v.13D no.7 s.110
• /
• pp.1027-1038
• /
• 2006

There are many technical problems in the recommendation system based on very large database(VLDB). So, it is necessary to study the recommendation system' structure and the data-mining technique suitable for the large scale Internet shopping mail. Thus we design and implement the product recommendation system using k-means clustering algorithm and sequential pattern technique which can be used in large scale Internet shopping mall. This paper processes user information by batch processing, defines the various categories by hierarchical structure, and uses a sequential pattern mining technique for the search engine. For predictive modeling and experiment, we use the real data(user's interest and preference of given category) extracted from log file of the major Internet shopping mall in Korea during 30 days. And we define PRP(Predictive Recommend Precision), PRR(Predictive Recommend Recall), and PF1(Predictive Factor One-measure) for evaluation. In the result of experiments, the best recommendation time and the best learning time of our system are much as O(N) and the values of measures are very excellent.