• ETRI Journal
• /
• 제46권1호
• /
• pp.118-126
• /
• 2024

In human conversations, listeners often utilize brief backchannels such as "uh-huh" or "yeah." Timely backchannels are crucial to understanding and increasing trust among conversational partners. In human-machine conversation systems, users can engage in natural conversations when a conversational agent generates backchannels like a human listener. We propose a method that simultaneously predicts backchannels and recognizes speech in real time. We use a streaming transformer and adopt multitask learning for concurrent backchannel prediction and speech recognition. The experimental results demonstrate the superior performance of our method compared with previous works while maintaining a similar single-task speech recognition performance. Owing to the extremely imbalanced training data distribution, the single-task backchannel prediction model fails to predict any of the backchannel categories, and the proposed multitask approach substantially enhances the backchannel prediction performance. Notably, in the streaming prediction scenario, the performance of backchannel prediction improves by up to 18.7% compared with existing methods.

• 한국지능시스템학회논문지
• /
• 제10권6호
• /
• pp.616-628
• /
• 2000

In this paper, we present the modeling of the ozone prediction system using Neuro-Fuzzy approaches. The mechanism of ozone concentration is highly complex, nonlinear, and nonstationary, the modeling of ozone prediction system has many problems and the results of prediction is not a good performance so far. The Dynamic Polynomial Neural Network(DPNN) which employs a typical algorithm of GMDH(Group Method of Data Handling) is a useful method for data analysis, identification of nonlinear complex system, and prediction of a dynamical system. The structure of the final model is compact and the computation speed to produce an output is faster than other modeling methods. In addition to DPNN, this paper also includes a Fuzzy Logic Method for modeling of ozone prediction system. The results of each modeling method and the performance of ozone prediction are presented. The proposed method shows that the prediction to the ozone concentration based upon Neuro-Fuzzy approaches gives us a good performance for ozone prediction in high and low ozone concentration with the ability of superior data approximation and self organization.

• 한국유체기계학회 논문집
• /
• 제6권2호
• /
• pp.29-33
• /
• 2003

The periphery pump (or regenerative pump) has been generally applied in the automotive fuel pump due to their low specific speed (high heads and small flow rate) with stable performance curves. In this study, the performance prediction of side channel type periphery pumps has been developed. The prediction of the circulatory flow rate is based on the consideration of the centrifugal force field in the side-channel and in the impeller vane grooves. For the determination of performance curve (head-flow rate), momentum exchange theory is used. The effects of various geometric parameters and loss coefficients used in the performance prediction method on the head and efficiency are discussed, and the results were compared with experimental data.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2002년도 학술대회지
• /
• pp.581-584
• /
• 2002

The periphery pump(or regenerative pump) has been generally applied in the automotive fuel pump due to their low specific speed(high heads and small flow rate) with stable performance curves. In this study, the performance prediction of side channel type periphery pumps has been developed. The prediction of the circulatory flow rate is based on the consideration of the centrifugal force field in the side-channel and in the impeller vane grooves. For the determination of performance curve(head-flow rate), momentum exchange theory is used. The effects of various geometric parameters and loss coefficients used in the performance prediction method on the head and efficiency are discussed and the results were compared with experimental data.

• 한국정보시스템학회지:정보시스템연구
• /
• 제30권4호
• /
• pp.247-275
• /
• 2021

Purpose This study attempted to predict corporate innovation and business performance using ensemble learning. Design/methodology/approach The ensemble techniques uses weak learning to create robust learning, which combines several weak models to derive improved performance. In this study, XGboost, LightGBM, and Catboost were used among ensemble techniques. It was compared and evaluated with traditional machine learning methods. Findings The summary of the research results is as follows. First, the type of innovation is expanding from technical innovation to non-technical areas. Second, it was confirmed that LightGBM performed best for radical innovation prediction, and XGboost performed best for incremental innovation prediction. Third, Catboost performed best for firm performance prediction. Although there was no significant difference in predictive power between ensemble techniques, we found that comparative analysis was necessary to confirm better prediction performance.

• 유체기계공업학회:학술대회논문집
• /
• 유체기계공업학회 2005년도 연구개발 발표회 논문집
• /
• pp.378-383
• /
• 2005

The performances of side channel type regenerative blowers were evaluated by the blower performance test, 1-D performance prediction and CFD. The performance prediction method was modified using the results of the performance test and CFD and applied to the design of the new regenerative blowers. The major geometric parameters such as channel height, channel area and expansion angle were decided from the performance prediction method for the improved models and the predicted results were compared with CFD and experimental data. Both of the modified models showed improved efficiency at the operating condition. Especially, model3 could be possible to reduce operating rotating speed, that is benefit to noise performance, because of the high head performance at the design point. The CFD results showed that the performance of the regenerative blower was influenced by the secondary circulatory flow in the channel.

• 대기
• /
• 제29권1호
• /
• pp.21-39
• /
• 2019

The conventional observations of the Korea Meteorological Administration (KMA) and National Centers for Environmental Prediction (NCEP) are compared in the numerical weather forecast system at the Korea Institute of Atmospheric Prediction Systems (KIAPS). The weather forecasting system used in this study is consists of Korea Integrated Model (KIM) as a global numerical weather prediction model, three-dimensional variational method as a data assimilation system, and KIAPS Package for Observation Processing (KPOP) as an observation pre-processing system. As a result, the forecast performance of NCEP observation was better while the number of observation is similar to the KMA observation. In addition, the sensitivity of forecast performance was investigated for each SONDE, SURFACE and AIRCRAFT observations. The differences in AIRCRAFT observation were not sensitive to forecast, but the use of NCEP SONDE and SURFACE observations have shown better forecast performance. It is found that the NCEP observations have more wind observations of the SONDE in the upper atmosphere and more surface pressure observations of the SURFACE in the ocean. The results suggest that evenly distributed observations can lead to improved forecast performance.

• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제16권11호
• /
• pp.1041-1045
• /
• 2010

클러스터는 병렬 컴퓨팅 및 데이터 센터에 적합한 구조를 제시하지만 설계 빛 확장을 할 때 성능에 대한 예측이 쉽지 않다 또한 기존의 클러스터 성능 분석은 이미 구성된 시스템만을 그 대상으로 한다는 문제점을 가지고 있으며 클러스터의 확장 및 대용량 클러스터에 대한 성능 예측을 지원하지 못한다. 그러므로 기존에 대규모 클러스터를 평가하던 방법들과는 다른, 시스템 구성 전 대규모 클러스터를 위한 모델링 및 예측 방법을 필요로 한다. 이러한 작업은 클러스터의 구조적 특성이 잘 반영되어야 하며 실제 시스템 적용 시 나타나는 문제에 관해서도 분석이 쉽게 되어야 한다. 본 논문에서는 대규모 클러스터의 성능 모델링을 위한 방법론을 제시하고 실제 시스템에서 수행한 측정 및 예측 결과로 방법론의 유용성을 보인다.

• KEPCO Journal on Electric Power and Energy
• /
• 제6권4호
• /
• pp.413-418
• /
• 2020

초기 학습 데이터의 과적합으로 인한 전력망 상태예측 모델의 성능 감소를 방지하고 예측모델의 예측 정확도 유지를 통한 계속적인 현장활용을 위해서는 기계학습 모델의 예측 정확도를 지속적으로 관리할 필요가 있다. 이를 위해, 본 논문에서는 다양한 요인에 의해 끊임없이 변화하는 전력망 상태 데이터의 특성을 고려하여 예측모델의 정확성과 신뢰성을 높이고 현장 적용 가능한 수준의 품질을 유지하기 위한 기계학습 기반 전력망 상태예측 모델의 성능 유지관리 자동화 기법을 제안한다. 제안 기법은 워크플로우 관리 기술의 적용을 통해 전력망 상태예측 모델 성능 유지관리를 위한 일련의 태스크들을 워크플로우의 형태로 모델링하고 이를 자동화하여 업무를 효율화 하였다. 또한, 기존 기술에서는 시도되지 않았던 학습데이터의 통계적 특성 변화 정도와 예측의 일반화 수준을 모두 고려한 예측모델의 성능 평가를 통해 성능 결과의 신뢰성을 확보하고 이를 통해 예측 모델의 정확도를 일정 수준으로 유지관리하고 더욱 성능이 우수한 예측모델의 신규 개발이 가능하다. 결과적으로 본 논문에서 제안하는 전력망 상태예측 모델 성능 유지관리 자동화 기법을 통해 예측모델의 성능 저하문제를 해결하여 분산자원 연계 등 외부 환경의 변화에 유연한 예측모델 관리를 통해 정확성과 신뢰성이 보장된 예측 모델의 지속적인 활용이 가능하다.

• 한국재료학회지
• /
• 제33권4호
• /
• pp.164-174
• /
• 2023

The bandgap characteristics of semiconductor materials are an important factor when utilizing semiconductor materials for various applications. In this study, based on data provided by AFLOW (Automatic-FLOW for Materials Discovery), the bandgap of a semiconductor material was predicted using only the material's compositional features. The compositional features were generated using the python module of 'Pymatgen' and 'Matminer'. Pearson's correlation coefficients (PCC) between the compositional features were calculated and those with a correlation coefficient value larger than 0.95 were removed in order to avoid overfitting. The bandgap prediction performance was compared using the metrics of R² score and root-mean-squared error. By predicting the bandgap with randomforest and xgboost as representatives of the ensemble algorithm, it was found that xgboost gave better results after cross-validation and hyper-parameter tuning. To investigate the effect of compositional feature selection on the bandgap prediction of the machine learning model, the prediction performance was studied according to the number of features based on feature importance methods. It was found that there were no significant changes in prediction performance beyond the appropriate feature. Furthermore, artificial neural networks were employed to compare the prediction performance by adjusting the number of features guided by the PCC values, resulting in the best R² score of 0.811. By comparing and analyzing the bandgap distribution and prediction performance according to the material group containing specific elements (F, N, Yb, Eu, Zn, B, Si, Ge, Fe Al), various information for material design was obtained.