• 한국분무공학회지
• /
• 제25권4호
• /
• pp.188-195
• /
• 2020

To meet stringent emission regulations of automotive engines, fuel injection control techniques have advanced based on reliable and fast computing prediction models. This study aims to develop a reliable lightweight prediction model of fuel injection rates using a small number of input parameters and based on simple fluid dynamic theories. The prediction model uses the geometry of the injector nozzle, needle motion data, injection conditions and the fuel properties. A commercial diesel injector and US No. 2 diesel were used as the test injector and fuel, respectively. The needle motion data were measured using X-ray phase-contrast imaging technique under various fuel injection pressures and injection pulse durations. The actual injector rate profiles were measured using an injection rate meter for the validation of the model prediction results. In the case of long injection durations with the steady-state operation, the model prediction results showed over 99 % consistency with the measurement results. However, in the case of short injection cases with the transient operation, the prediction model overestimated the injection rate that needs to be further improved.

• 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.

• Interdisciplinary Bio Central
• /
• 제4권4호
• /
• pp.14.1-14.6
• /
• 2012

Splice site prediction in DNA sequence is a basic search problem for finding exon/intron and intron/exon boundaries. Removing introns and then joining the exons together forms the mRNA sequence. These sequences are the input of the translation process. It is a necessary step in the central dogma of molecular biology. The main task of splice site prediction is to find out the exact GT and AG ended sequences. Then it identifies the true and false GT and AG ended sequences among those candidate sequences. In this paper, we survey research works on splice site prediction based on support vector machine (SVM). The basic difference between these research works is nucleotide encoding technique and SVM kernel selection. Some methods encode the DNA sequence in a sparse way whereas others encode in a probabilistic manner. The encoded sequences serve as input of SVM. The task of SVM is to classify them using its learning model. The accuracy of classification largely depends on the proper kernel selection for sequence data as well as a selection of kernel parameter. We observe each encoding technique and classify them according to their similarity. Then we discuss about kernel and their parameter selection. Our survey paper provides a basic understanding of encoding approaches and proper kernel selection of SVM for splice site prediction.

• 대한전자공학회논문지SD
• /
• 제45권9호
• /
• pp.103-109
• /
• 2008

본 논문은 H.264 디코더용 인트라 예측 회로 구조와 설계를 제시한다. H.264의 인트라 예측에는 총 17개의 예측 모드, 즉 루마 $4\times4$ 블록을 위한 9개의 예측 모드, 루마 $16\times16$ 블록을 위한 4개의 예측 모드, 크로마 $8\times8$ 블록을 위한 4개의 예측 모드가 있다 모든 예측 모드에서 공통된 연산들을 추출하여 이들을 수행하기 위한 공통연산부를 정의하였다. 모든 예측 모드에서 이 연산부를 공유하는 제안된 회로 구조는 설계 측면에서 체계적이고 회로 크기 측면에서 효율적이다.

• 대기
• /
• 제25권2호
• /
• pp.323-337
• /
• 2015

This study explores the 6-month lead prediction skill of several climate indices that influence on East Asian climate in the GloSea5 hindcast experiment. Such indices include Nino3.4, Indian Ocean Diploe (IOD), Arctic Oscillation (AO), various summer and winter Asian monsoon indices. The model's prediction skill of these indices is evaluated by computing the anomaly correlation coefficient (ACC) and mean squared skill score (MSSS) for ensemble mean values over the period of 1996~2009. In general, climate indices that have low seasonal variability are predicted well. For example, in terms of ACC, Nino3.4 index is predicted well at least 6 months in advance. The IOD index is also well predicted in late summer and autumn. This contrasts with the prediction skill of AO index which shows essentially no skill beyond a few months except in February and August. Both summer and winter Asian monsoon indices are also poorly predicted. An exception is the Western North Pacific Monsoon (WNPM) index that exhibits a prediction skill up to 4- to 6-month lead time. However, when MSSS is considered, most climate indices, except Nino3.4 index, show a negligible prediction skill, indicating that conditional bias is significant in the model. These results are only weakly sensitive to the number of ensemble members.

• ETRI Journal
• /
• 제27권5호
• /
• pp.511-524
• /
• 2005

In this paper, we propose a novel hardware architecture for an intra-prediction, integer transform, quantization, inverse integer transform, inverse quantization, and mode decision module for the macroblock engine of a new video coding standard, H.264. To reduce the cycle of intra prediction, transform/quantization, and inverse quantization/inverse transform of H.264, a reduction method for cycle overhead in the case of I16MB mode is proposed. This method can process one macroblock for 927 cycles for all cases of macroblock type by processing $4{\times}4$ Hadamard transform and quantization during $16{\times}16$ prediction. This module was designed using Verilog Hardware Description Language (HDL) and operates with a 54 MHz clock using the Hynix $0.35 {\mu}m$ TLM (triple layer metal) library.

• Genomics & Informatics
• /
• 제15권4호
• /
• pp.162-169
• /
• 2017

Metal binding proteins or metallo-proteins are important for the stability of the protein and also serve as co-factors in various functions like controlling metabolism, regulating signal transport, and metal homeostasis. In structural genomics, prediction of metal binding proteins help in the selection of suitable growth medium for overexpression's studies and also help in obtaining the functional protein. Computational prediction using machine learning approach has been widely used in various fields of bioinformatics based on the fact all the information contains in amino acid sequence. In this study, random forest machine learning prediction systems were deployed with simplified amino acid for prediction of individual major metal ion binding sites like copper, calcium, cobalt, iron, magnesium, manganese, nickel, and zinc.

• Genomics & Informatics
• /
• 제14권4호
• /
• pp.149-159
• /
• 2016

With the success of the genome-wide association studies (GWASs), many candidate loci for complex human diseases have been reported in the GWAS catalog. Recently, many disease prediction models based on penalized regression or statistical learning methods were proposed using candidate causal variants from significant single-nucleotide polymorphisms of GWASs. However, there have been only a few systematic studies comparing existing methods. In this study, we first constructed risk prediction models, such as stepwise linear regression (SLR), least absolute shrinkage and selection operator (LASSO), and Elastic-Net (EN), using a GWAS chip and GWAS catalog. We then compared the prediction accuracy by calculating the mean square error (MSE) value on data from the Korea Association Resource (KARE) with body mass index. Our results show that SLR provides a smaller MSE value than the other methods, while the numbers of selected variables in each model were similar.

• 한국항공운항학회지
• /
• 제24권4호
• /
• pp.44-52
• /
• 2016

The departure flow management is the planning tool to optimize the schedule of the departure aircraft and allows them to join smoothly into the overhead traffic flow. To that end, the arrival time prediction to the merge point for the cruising aircraft is necessary to determined. This paper proposes a trajectory prediction model for the cruising aircraft based on the machine learning approach. The proposed method includes the trajectory vectored from the procedural route and is applied to the historical data to evaluate the prediction performances.

• 정보통신설비학회논문지
• /
• 제10권2호
• /
• pp.74-79
• /
• 2011

The goal of this research is to reduce dynamic and static power consumption for a low power cache system. The proposed cache can achieve a low power consumption by using a drowsy and a way prediction mechanism. For reducing the static power, the drowsy technique is used at 4-way set associative cache. And for reducing the dynamic energy, one among four ways is selectively accessed on the basis of information in the Way-Line Prediction Unit (WLPU). This prediction mechanism does not introduce any additional delay though prediction misses are occurred. The WLPU can effectively reduce the performance overhead of the conventional drowsy caching by waking only a drowsy cache line and one way in advance. Our results show that the proposed cache can reduce the power consumption by about 40% compared with the 4-way drowsy cache.