• Animal Bioscience
• /
• v.36 no.8
• /
• pp.1199-1208
• /
• 2023

Objective: The purpose of this study was to perform a comparative analysis of metabolite levels in serum and milk obtained from cows fed on different concentrate to forage feed ratios. Methods: Eight lactating Holstein cows were divided into two groups: a high forage ratio diet (HF; 80% Italian ryegrass and 20% concentrate of daily intake of dry matter) group and a high concentrate diet (HC; 20% Italian ryegrass and 80% concentrate) group. Blood was collected from the jugular vein, and milk was sampled using a milking machine. Metabolite levels in serum and milk were estimated using proton nuclear magnetic resonance and subjected to qualitative and quantitative analyses performed using Chenomx 8.4. For statistical analysis, Student's t-test and multivariate analysis were performed using Metaboanalyst 4.0. Results: In the principal component analysis, a clear distinction between the two groups regarding milk metabolites while serum metabolites were shown in similar. In serum, 95 metabolites were identified, and 13 metabolites (include leucine, lactulose, glucose, betaine, etc.) showed significant differences between the two groups. In milk, 122 metabolites were identified, and 20 metabolites (include urea, carnitine, acetate, butyrate, arabinitol, etc.) showed significant differences. Conclusion: Our results show that different concentrate to forage feed ratios impact the metabolite levels in the serum and milk of lactating Holstein cows. A higher number of metabolites in milk, including those associated with milk fat synthesis and the presence of Escherichia coli in the rumen, differed between the two groups compared to that in the serum. The results of this study provide a useful insight into the metabolites associated with different concentrate to forge feed ratios in cows and may aid in the search for potential biomarkers for subacute ruminal acidosis.

• Journal of Forest and Environmental Science
• /
• v.9 no.1
• /
• pp.57-66
• /
• 1993

The purpose of this study is to offer detailed topographic information for substantially selecting the yarding machine for mechanized yarding operations, classifying the forest site by cluster analysis and principal component analysis, and investigating simultaneously the variables which give much influence on the classification of forest site in forestry build-up region (21, 477ha) of Chunchon Gun, Kwangweon Do. Ten topographic variables were used for the analysis. The results of study were as follows : 1) Gosung region (2, 252ha) was classified into hilly terrain (57%) and steep terrain (43%) and required the tractor prehauling system for the former one and the medium skyline system for latter one, respectively. 2) 65% of Gajung region (2,306ha) and 67% of Kwangpan region (2, 627ha) were classified into steep terrain fitted for the medium skyline system and the portion of both region showed the hilly terrain for the tractor prehauling system. 3) Jiam region (4, 591ha), consisted only of steep terrain, required the medium skyline system. 4) Gunja region (3, 400ha), Sudong region (3, 984ha) and Sinpo region (2, 340ha) were classified into steep terrain, requiring the medium skyline system, with 85%, 75%, and 75%, respectively.

• Current Optics and Photonics
• /
• v.4 no.1
• /
• pp.31-43
• /
• 2020

The feasibility of the application of terahertz electromagnetic waves in the diagnosis of prostate cancer was examined. Four samples of incomplete cancerous prostatic paraffin-embedded tissues were examined using terahertz spectral imaging (TPI) system and the results obtained by comparing the absorption coefficient and refractive index of prostate tumor, normal prostate tissue and smooth muscle from one of the paraffin tissue masses examined were reported. Three hundred and sixty cases of absorption coefficients from one of the paraffin tissues examined were used as raw data to classify these three tissues using the Principal Component Analysis (PCA) and Least Squares Support Vector Machine (LS-SVM). An excellent classification with an accuracy of 92.22% in the prediction set was achieved. Using the distribution information of THz reflection signal intensity from sample surface and absorption coefficient of the sample, an attempt was made to use the TPI system to identify the boundaries of the different tissues involved (prostate tumors, normal and smooth muscles). The location of three identified regions in the terahertz images (frequency domain slice absorption coefficient imaging, 1.2 THz) were compared with those obtained from the histopathologic examination. The tissue tumor region had a distinctively visible color and could well be distinguished from other tissue regions in terahertz images. Results indicate that a THz spectroscopy imaging system can be efficiently used in conjunction with the proposed advanced computer-based mathematical analysis method to identify tumor regions in the paraffin tissue mass of prostate cancer.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.11
• /
• pp.527-536
• /
• 2017

With the recent development of smart grid industry, the necessity for efficient EMS(Energy Management System) has been increased. In particular, in order to reduce electric load and energy cost, sophisticated electric load forecasting and efficient smart grid operation strategy are required. In this paper, for more accurate electric load forecasting, we extend the data collected at demand time into high time resolution and construct an artificial neural network-based forecasting model appropriate for the high time resolution data. Furthermore, to improve the accuracy of electric load forecasting, time series data of sequence form are transformed into continuous data of two-dimensional space to solve that problem that machine learning methods cannot reflect the periodicity of time series data. In addition, to consider external factors such as temperature and humidity in accordance with the time resolution, we estimate their value at the time resolution using linear interpolation method. Finally, we apply the PCA(Principal Component Analysis) algorithm to the feature vector composed of external factors to remove data which have little correlation with the power data. Finally, we perform the evaluation of our model through 5-fold cross-validation. The results show that forecasting based on higher time resolution improve the accuracy and the best error rate of 3.71% was achieved at the 3-min resolution.

• Journal of KIISE
• /
• v.42 no.12
• /
• pp.1535-1543
• /
• 2015

The brain-computer interface obtains a user's electroencephalogram as a replacement communication unit for the disabled such that the user is able to control machines by simply thinking instead of using hands or feet. In this paper, we propose a feature extraction method based on a non-selected filter by SBCSP to classify motor imagery EEG. First, we divide frequencies (4~40 Hz) into 4-Hz units and apply CSP to each Unit. Second, we obtain the FLD score vector by combining FLD results. Finally, the FLD score vector is projected onto the optimal plane for classification using PCA. We use BCI Competition III dataset IVa, and Extracted features are used as input for LS-SVM. The classification accuracy of the proposed method was evaluated using $10{\times}10$ fold cross-validation. For subjects 'aa', 'al', 'av', 'aw', and 'ay', results were $85.29{\pm}0.93%$, $95.43{\pm}0.57%$, $72.57{\pm}2.37%$, $91.82{\pm}1.38%$, and $93.50{\pm}0.69%$, respectively.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.22 no.3
• /
• pp.471-479
• /
• 2018

Recently, the spread of smart TV based Android iOS Set Top box has become common. This paper propose a new approach to control the TV using gestures away from the era of controlling the TV using remote control. In this paper, the AdaBoost algorithm is applied to gesture recognition by using a mono camera. First, we use Camshift-based Body tracking and estimation algorithm based on Gaussian background removal for body coordinate extraction. Using global and local feature vectors, we recognized gestures with speed change. By tracking the time interval trajectories of hand and wrist, the AdaBoost algorithm with CART algorithm is used to train and classify gestures. The principal component feature vector with high classification success rate is searched using CART algorithm. As a result, 24 optimal feature vectors were found, which showed lower error rate (3.73%) and higher accuracy rate (95.17%) than the existing algorithm.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.38 no.2
• /
• pp.227-233
• /
• 2014

This study investigates the vibration characteristics of a wire-bonding piezoelectric transducer and ultrasonic horn for high-speed and precise welding. A ring-type piezoelectric stack actuator is excited at 136 kHz to vibrate a conical-type horn and capillary system. The nodal lines and amplification ratio of the ultrasonic horn are obtained using a theoretical analysis and FEM simulation. The vibration modes and frequencies close to the driving frequency are identified to evaluate the bonding performance of the current wire-bonder system. The FEM and experimental results show that the current wire-bonder system uses the bending mode of 136 kHz as the principal motion for bonding and that the transverse vibration of the capillary causes the bonding failure. Because the major longitudinal mode exists at 119 kHz, it is recommended that the design of the current wire-bonding system be modified to use the major longitudinal mode at the excitation frequency and to minimize the transverse vibration of capillary in order to improve the bonding performance.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.10
• /
• pp.465-472
• /
• 2017

In the past, researchers mainly used the supervised learning technique of machine learning to analyze power data and investigated the identification of patterns through the data mining technique. Data analysis research, however, faces its limitations with the old data classification and analysis techniques today when the size of electric power data has increased with the possible real-time provision of data. This study thus set out to propose a clustering architecture to analyze large-sized electric power data. The clustering process proposed in the study supplements the K-means algorithm, an unsupervised learning technique, for its problems and is capable of automating the entire process from the collection of electric power data to their analysis. In the present study, power data were categorized and analyzed in total three levels, which include the row data level, clustering level, and user interface level. In addition, the investigator identified K, the ideal number of clusters, based on principal component analysis and normal distribution and proposed an altered K-means algorithm to reduce data that would be categorized as ideal points in order to increase the efficiency of clustering.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.164-175
• /
• 2019

The Tor Tong Daeng Irrigation Project with the irrigation area of 61,400 hectares is located in the Ping Basin of the Upper Central Plain of Thailand where farmers depended on both surface water and groundwater. In the drought year, water storage in the Bhumipol Dam is inadequate to allocate water for agriculture, and caused water deficit in many irrigation projects. Farmers need to find extra sources of water such as water from farm pond or groundwater as a supplement. The operation of Bhumipol Dam and irrigation demand estimation are vital for irrigation water allocation to help solve water shortage issue in the irrigation project. The study aims to determine the smart dam operation system to mitigate water shortage in this irrigation project via introduction of machine learning to improve dam operation and irrigation demand estimation via soil moisture estimation from satellite images. Via ANN technique application, the inflows to the dam are generated from the upstream rain gauge stations using past 10 years daily rainfall data. The input vectors for ANN model are identified base on regression and principal component analysis. The structure of ANN (length of training data, the type of activation functions, the number of hidden nodes and training methods) is determined from the statistics performance between measurements and ANN outputs. On the other hands, the irrigation demand will be estimated by using satellite images, LANDSAT. The Enhanced Vegetation Index (EVI) and Temperature Vegetation Dryness Index (TVDI) values are estimated from the plant growth stage and soil moisture. The values are calibrated and verified with the field plant growth stages and soil moisture data in the year 2017-2018. The irrigation demand in the irrigation project is then estimated from the plant growth stage and soil moisture in the area. With the estimated dam inflow and irrigation demand, the dam operation will manage the water release in the better manner compared with the past operational data. The results show how smart system concept was applied and improve dam operation by using inflow estimation from ANN technique combining with irrigation demand estimation from satellite images when compared with the past operation data which is an initial step to develop the smart dam operation system in Thailand.

• Journal of Broadcast Engineering
• /
• v.27 no.1
• /
• pp.56-68
• /
• 2022

In this paper, we propose a compression method for multi-resolution feature maps for VCM. The proposed compression method removes the redundancy between the channels and resolution levels of the multi-resolution feature map through PCA-based transformation. According to each characteristic, the basis vectors and mean vector used for transformation, and the transformation coefficient obtained through the transformation are compressed using a VVC-based coder and DeepCABAC. In order to evaluate performance of the proposed method, the object detection performance was measured for the OpenImageV6 and COCO 2017 validation set, and the BD-rate of MPEG-VCM anchor and feature map compression anchor proposed in this paper was compared using bpp and mAP. As a result of the experiment, the proposed method shows a 25.71% BD-rate performance improvement compared to feature map compression anchor in OpenImageV6. Furthermore, for large objects of the COCO 2017 validation set, the BD-rate performance is improved by up to 43.72% compared to the MPEG-VCM anchor.