• 한국컴퓨터정보학회논문지
• /
• 제1권1호
• /
• pp.129-138
• /
• 1996

산업발전과 기술의 대형화. 고도화 등으로 인하여 매년 방대한 양리 정보가 처리되고 있다 정보화를 이루기 위해서는 대부분 종이로 기록뇌어 내려오던 정보를 컴퓨터에 저장하여 적기적소에 사용할 수 있어야 한다. 문자인식을 위한 신경망의 학습에 있어서 출력 값을 재사용하는 신경망모델로는 순환신경망이 있다. 그러나 이러한 방법들의 대부분은 오프라인 필기체문자와 같은 정적인 패턴의 분류에 있어서는 효과적으로 적락되지 않는다. 이에 본 연구에서는 오프라인 필기체문자와 같은 정적인 패턴을 효과적으로 분류하기 위한 새로운 형태의 순환신경망을 제안한다. 본 논문은 Jordan과 Elman Model을 확장 결합한 새로운 J-도(Jordan-Elman) 신경망 모델을 사용하여 숫자 및 필기체 문자와 같은 정적인 패턴의 인식에서 기존의 신경망보다 성능이 향상되었음을 보여준다.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권11호
• /
• pp.1464-1471
• /
• 1999

This paper introduces complex-valued competitive learning neural network for spatio-temporal pattern recognition. There have been quite a few neural networks for spatio-temporal pattern recognition. Among them, recurrent neural network, TDNN, and avalanche model are acknowledged as standard neural network paradigms for spatio-temporal pattern recognition. Recurrent neural network has complicated learning rules and does not guarantee convergence to global minima. TDNN requires too many neurons, and can not be regarded to deal with spatio-temporal pattern basically. Grossberg's avalanche model is not able to distinguish long patterns, and has to be indicated which layer is to be used in learning. In order to remedy drawbacks of the above networks, unsupervised competitive learning using complex umber is proposed. Suggested neural network also features simultaneous recognition, time-shift invariant recognition, stable categorizing, and learning rate modulation. The network is evaluated by computer simulation with randomly generated patterns.

• 한국통신학회논문지
• /
• 제21권2호
• /
• pp.534-540
• /
• 1996

This paper presents an Adaptive Time-Delay Recurrent Neural Network (ATRN) for learning and recognition of temporal correlations of temporal patterns. The ATRN employs adaptive time-delays and recurrent connections, which are inspired from neurobiology. In the ATRN, the adaptive time-delays make the ATRN choose the optimal values of time-delays for the temporal location of the important information in the input parrerns, and the recurrent connections enable the network to encode and integrate temporal information of sequences which have arbitrary interval time and arbitrary length of temporal context. The ATRN described in this paper, ATNN proposed by Lin, and TDNN introduced by Waibel were simulated and applied to the chaotic time series preditcion of Mackey-Glass delay-differential equation. The simulation results show that the normalized mean square error (NMSE) of ATRN is 0.0026, while the NMSE values of ATNN and TDNN are 0.014, 0.0117, respectively, and in temporal learning, employing recurrent links in the network is more effective than putting multiple time-delays into the neurons. The best performance is attained bythe ATRN. This ATRN will be sell applicable for temporally continuous domains, such as speech recognition, moving object recognition, motor control, and time-series prediction.

• Childhood Kidney Diseases
• /
• 제28권1호
• /
• pp.35-43
• /
• 2024

Purpose: Recurrent urinary tract infections (UTIs) in children is a major challenge for pediatricians. This study was designed to investigate the risk factors for recurrent UTIs and determine the association between recurrent UTIs and clinical findings, including growth patterns in infants and children younger than 24 months of age. Methods: We retrospectively reviewed the medical records of 147 patients <24 months of age with UTIs who were hospitalized between August 2018 and October 2021. The patients were divided into recurrent and single UTI episode groups. Clinical findings and anthropometric and laboratory data were compared between the two groups. Results: In the recurrent UTI group, the weight-for-length (WFL) percentile at the first UTI diagnosis was lower compared to the single UTI episode group, and the weight-for-age percentile at 3-month and 6-month follow-ups after the first UTI decreased (all P<0.05). In univariable logistic regression analysis, higher birth weight, lower WFL percentile, the presence of hydronephrosis, acute pyelonephritis or vesicoureteral reflux, the use of prophylactic antibiotics, and non-Escherichia coli infections were associated with the development of recurrent UTIs (all P<0.05). However, in the multivariable analysis, only the presence of hydronephrosis and prophylactic antibiotic use were independently related to UTI recurrence (P<0.05). Conclusions: The presence of hydronephrosis at the first UTI can be helpful for predicting UTI recurrence in young children aged <24 months. Antibiotic prophylaxis may be associated with UTI recurrence. Potential growth delay should be carefully monitored in infants with recurrent UTI.

• 대한교통학회지
• /
• 제32권5호
• /
• pp.522-530
• /
• 2014

교통관리센터에 축적되어 있는 속도 이력데이터에는 반복 비반복 정체 시공간 전개에 대한 상세한 정보가 모두 들어있으나, 도해법에 의해 다루어져 왔기 때문에 많은 양의 이력데이터를 처리하여 교통상황예측이나 정보제공에 활용할 수 없는 한계가 존재하였다. 본 논문에서는, 기존의 Classification과 Density-Based Clustering 알고리즘을 속도 시공간 데이터 특성에 맞게 조합하고 변형하여 정체 시공간 영역을 자동 인식하는 알고리즘과, 정체파급길이, 파급속도, 해소속도 등 정체 시공간 전개 패턴의 특성치를 산정하는 알고리즘을 개발하였다, 본 알고리즘은, 교통관리센터에 축적되어 있는 방대한 양의 이력데이터를 자동으로 분석하여 자세한 정체 관련 정보를 추출할 수 있고, 산정된 특성치를 가지고 각 센터의 필요에 따라 다양한 정보를 2차 생성하고 활용할 수 있는 장점이 있다. 본 연구결과는 향후 반복 비반복 정체에 대한 예측과 대응이 획기적으로 개선되는데 초석이 될 것으로 기대된다.

• Journal of Gastric Cancer
• /
• 제22권4호
• /
• pp.369-380
• /
• 2022

Purpose: Tumor recurrence is the principal cause of poor outcomes in remnant gastric cancer (RGC) after resection. We sought to elucidate the recurrent patterns according to tumor locations in RGC. Materials and Methods: Data were collected from the Shanghai Cancer Center between January 2006 and December 2020. A total of 129 patients with RGC were included in this study, of whom 62 had carcinomas at the anastomotic site (group A) and 67 at the non-anastomotic site (group N). The clinicopathological characteristics, surgical results, recurrent diseases, and survival were investigated according to tumor location. Results: The time interval from the previous gastrectomy to the current diagnosis was 32.0±13.0 and 21.0±13.4 years in groups A and N, respectively. The previous disease was benign in 51/62 cases (82.3%) in group A and 37/67 cases (55.2%) in group N (P=0.002). Thirty-three patients had documented sites of tumor recurrence through imaging or pathological examinations. The median time to recurrence was 11.0 months (range, 1.0-35.1 months). Peritoneal recurrence occurred in 11.3% (7/62) of the patients in group A versus 1.5% (1/67) of the patients in group N (P=0.006). Hepatic recurrence occurred in 3.2% (2/62) of the patients in group A versus 13.4% (9/67) of the patients in group N (P=0.038). Patients in group A had significantly better overall survival than those in group N (P=0.046). Conclusions: The tumor location of RGC is an essential factor for predicting recurrence patterns and overall survival. When selecting an optimal postoperative follow-up program for RGC, physicians should consider recurrent features according to the tumor location.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권2호
• /
• pp.538-561
• /
• 2020

Recommender Systems (RecSys) have a major role in e-commerce for recommending products, which they may like for every user and thus improve their business aspects. Although many types of RecSyss are there in the research field, the state of the art RecSys has focused on finding the user similarity based on sequence (e.g. purchase history, movie-watching history) analyzing and prediction techniques like Recurrent Neural Network in Deep learning. That is RecSys has considered as a sequence prediction problem. However, evaluation of similarities among the customers is challenging while considering temporal aspects, context and multi-component ratings of the item-records in the customer sequences. For addressing this issue, we are proposing a Deep Learning based model which learns customer similarity directly from the sequence to sequence similarity as well as item to item similarity by considering all features of the item, contexts, and rating components using Dynamic Temporal Warping(DTW) distance measure for dynamic temporal matching and 2D-GRU (Two Dimensional-Gated Recurrent Unit) architecture. This will overcome the limitation of non-linearity in the time dimension while measuring the similarity, and the find patterns more accurately and speedily from temporal and spatial contexts. Experiment on the real world movie data set LDOS-CoMoDa demonstrates the efficacy and promising utility of the proposed personalized RecSys architecture.

• 반도체디스플레이기술학회지
• /
• 제23권1호
• /
• pp.1-6
• /
• 2024

Recently, artificial neural networks have been playing a crucial role and advancing across various fields. Artificial neural networks are typically categorized into feedforward neural networks and recurrent neural networks. However, feedforward neural networks are primarily used for processing static spatial patterns such as image recognition and object detection. They are not suitable for handling temporal signals. Recurrent neural networks, on the other hand, face the challenges of complex training procedures and requiring significant computational power. In this paper, we propose memristors suitable for an advanced form of recurrent neural networks called reservoir computing systems, utilizing a mask processor. Using the characteristic equations of Ti/TiOx/TaOy/Pt, Pt/TiOx/Pt, and Ag/ZnO-NW/Pt memristors, we generated current-voltage curves to verify their memristive behavior through the confirmation of hysteresis. Subsequently, we trained and inferred reservoir computing systems using these memristors with the NIST TI-46 database. Among these systems, the accuracy of the reservoir computing system based on Ti/TiOx/TaOy/Pt memristors reached 99%, confirming the Ti/TiOx/TaOy/Pt memristor structure's suitability for inferring speech recognition tasks.

• Radiation Oncology Journal
• /
• 제37권3호
• /
• pp.193-200
• /
• 2019

Purpose: To explore the role of salvage radiotherapy (RT) for recurrent thymoma as an alternative to surgery. Materials and Methods: Between 2007 and 2015, 47 patients who received salvage RT for recurrent thymoma at Yonsei Cancer Center were included in this study. Recurrent sites included initial tumor bed (n = 4), pleura (n = 19), lung parenchyma (n = 10), distant (n = 9), and multiple regions (n = 5). Three-dimensional conformal and intensity-modulated RT were used in 29 and 18 patients, respectively. Median prescribed dose to gross tumor was 52 Gy (range, 30 to 70 Gy), with equivalent doses in 2-Gy fractions (EQD₂). We investigated overall survival (OS), progression-free survival (PFS), and patterns of failure. Local failure after salvage RT was defined as recurrence at the target volume receiving >50% of the prescription dose. Results: Median follow-up time was 83 months (range, 8 to 299 months). Five-year OS and PFS were 70% and 22%, respectively. The overall response rate was 97.9%; complete response, 34%; partial response, 44.7%; and stable disease, 19.1%. In multivariate analysis, histologic type and salvage RT dose (≥52 Gy, EQD₂) were significantly associated with OS. The high dose group (≥52 Gy, EQD₂) had significantly better outcomes than the low dose group (5-year OS: 80% vs. 59%, p = 0.046; 5-year PFS: 30% vs. 14%, p=0.002). Treatment failure occurred in 34 patients; out-of-field failure was dominant (intra-thoracic recurrence 35.3%; extrathoracic recurrence 11.8%), while local failure rate was 5.8%. Conclusion: Salvage RT for recurrent thymoma using high doses and advanced precision techniques produced favorable outcomes, providing evidence that recurrent thymoma is radiosensitive.

• 대한두경부종양학회지
• /
• 제15권2호
• /
• pp.166-171
• /
• 1999

Background and Objectives: The aim of our study was to describe the appearance of recurrent and residual lesions in the head and neck tumors, and to evaluate the usefullness of CT and MRI. Materials and Methods: CT(n=42) and MRI(n=4) of 44 patients with recurrent head and neck tumors were reviewed retrospectively. Primary tumor sites were larynx/hypopharynx in 15, oral cavity/floor of mouth in 13, base of tongue/tonsil in 5, nasopharynx in 4, palate in 2, and others in 5 patients. Therapeutic modalities included sugery and radiotherapy in 23, radiotherapy in 11, surgery in 5, chemotherapy and radiotherapy in 4, and chemotherapy in 1 patient. Results: The patterns of tumor recurrence were nodal recurrence(n=17), primary tumor bed recurrence combined with nodal recurrence(n=12), primary tumor bed recurrence(n=10) and residual primary tumors(n=5). The most common appearance of residual/recurrent primary tumor on CT was focal or diffuse heterogenous mass with or without surrounding fat or muscle infiltration(25/27). On MRI, the recurrent lesions showed intermediate signal intensity on T1 weighted image and high signal intensity on T2 weighted image with heterogenous enhancement in the most cases(n=3). 38 out of 44 nodal recurrences(86%) which had been pathologically or clinically proved were more than 1 cm in diameter or contained central low density on CT scan. Conclusion: Although CT and MRI findings of recurrent and residual tumors of the head and neck were nonspecific, in the majority the lesions manifested as a mass at primary tumor bed and/or nodal disease including contralateral side of the neck. And CT and MRI are valuable for revealing above lesions.