• The Journal of Korean Society for Radiation Therapy
• /
• v.23 no.2
• /
• pp.91-96
• /
• 2011

Purpose: The aim of study is to expose a more uniform dose depending on the relationship between a body mass index in patients who underwent radiation therapy and an acquired dosimetric information by using a thermoluminescent dosimeter. Materials and Methods: Since 2006 to August 2011 we investigated 28 people who underwent radiation therapy were enrolled in AMC. Each patient was measured on the head, neck, chest, abdomen, pelvis, thigh, knee joint, and ankle joint using the thermoluminescent dosimeter. The measurement value of each points compared with the prescribed center point, abdominal point, and dose measurements of points on which to base the abdomen and the patient's body mass index (BMI) were compared with reference point, abdomen dose. Results: 28 patients on prescribed dose in the abdomen by which the center point, an average dose was $100.6{\pm}5.5%$, and the other seven measuring points with the average maximum difference among the head, neck, chest, pelvic, thigh, knee, and ankle were $92.8{\pm}4.2%$, $97.6{\pm}6.2%$, $96.4{\pm}5.5%$, $102.6{\pm}5.3%$, $103.4{\pm}7.9%$, $95.8{\pm}5.9%$, $96.1{\pm}5.5%$. The relationship of abdominal point dose and the patient's body mass index (BMI) was analyzed a scatter plot, and the result of linear relationship analysis by regression method, the regression of the dose (y) was -1.009 BMI (x) plus 123.3 and coefficient of determination ($R^2$) was represented 0.697. Conclusion: The total body irradiation treatment process was evaluated the dose deviation and then the prescribed dose by which the average abdominal dose was satisfied with $100.6{\pm}5.5%$. Results of the relationship analysis between BMI and dose, if we apply the correction value for each patients, it can be achieved more uniform dose delivery.

• Geotechnical Engineering
• /
• v.14 no.5
• /
• pp.89-110
• /
• 1998

Variations of backfill load on the metal-polyethylene composite( MPC ) pipes buried in various trenches backfill afterward were investigated in this paper. The backfill loads obtained by the finite element method( FBM ) were compared with those calculated by the well-known MarstonBpangler(M-5) theory. The reliability of the finite element analysis used in this study was examined by an inaitu best for the buried pipe. The backfill lords and deflections on the real-size pipe buried on-site were measured while increasing the backfill height. In addition, further investigations were made for the variations of the backfill loads as a function of several important parameters such as the backfill soil type, bach. height$(\leq4.0m)$, diameter of the pipe$(B.$1.0m)$, and trench width($\leq 3.0 B_c$). It is confirmed that the M-S theory predicts reasonably well the backfill loafs of the MPC of the M-S backfill coils be 0.13 and 0.15 for the SC and SM coils in the D unman soil model, respectively. The load ratio, Wu-s/WwgM for a narrow trench varies negligibly with the back(111 height but fiends to increase for a wide trench. The ratio increases with increasing diameter of the pipe for a narrow trench while decreasing for a wide trench. It is also found that the ratio generally decreases as the degree of compaction increases and BM soil exhibits larger load ratio than that of SC soil.

• Progress in Medical Physics
• /
• v.24 no.4
• /
• pp.253-258
• /
• 2013

The purpose of this study is to evaluate and analyze the relationship between the external radiation dose reconstruction which is transmitted from the patient who receives radiation treatment through electronic portal imaging device (EPID) and the internal dose derived from the Monte Carlo simulation. As a comparative analysis of the two cases, it is performed to provide a basic indicator for similar studies. The geometric information of the experiment and that of the radiation source were entered into Monte Carlo n-particle (MCNPX) which is the computer simulation tool and to derive the EPID images, a tally card in MCNPX was used for visualizing and the imaging of the dose information. We set to source to surface distance (SSD) 100 cm for internal measurement and EPID. And the water phantom was set to be 100 cm of the source to surface distance (SSD) for the internal measurement and EPID was set to 90 cm of SSD which is 10 cm below. The internal dose was collected from the water phantom by using mesh tally function in MCNPX, accumulated dose data was acquired by four-portal beam exposures. At the same time, after getting the dose which had been passed through water phantom, dose reconstruction was performed using back-projection method. In order to analyze about two cases, we compared the penetrated dose by calibration of itself with the absorbed one. We also evaluated the reconstructed dose using EPID and partially accumulated (overlapped) dose in water phantom by four-portal beam exposures. The sum dose data of two cases were calculated as each 3.4580 MeV/g (absorbed dose in water) and 3.4354 MeV/g (EPID reconstruction). The result of sum dose match from two cases shows good agreement with 0.6536% dose error.

• Journal of Soil and Groundwater Environment
• /
• v.14 no.3
• /
• pp.1-13
• /
• 2009

Alkalinity and total carbon contents were measured by acid neutralizing titration (ANT), back titration (BT), gravitational weighing (GW), non-dispersive infrared-total carbon (NDIR-TC) methods for assessing precision and accuracy of alkalinity and total carbon concentration in $CO_2$-rich water. Artificial $CO_2$-rich water(ACW: pH 6.3, alkalinity 68.8 meq/L, $HCO_3^-$ 2,235 mg/L) was used for comparing the measurements. When alkalinity measured in 0 hr, percent errors of all measurement were 0~12% and coefficient of variation were less than 4%. As the result of post-hoc analysis after repeated measure analysis of variance (RM-AMOVA), the differences between the pair of methods were not significant (within confidence level of 95%), which indicates that the alkalinity measured by any method could be accurate and precise when it measured just in time of sampling. In addition, alkalinity measured by ANT and NDIR-TC were not change after 24 and 48 hours open to atmosphere, which can be explained by conservative nature of alkalinity although $CO_2$ degas from ACW. On the other hand, alkalinity measured by BT and GW increased after 24 and 48 hours open to atmosphere, which was caused by relatively high concentration of measured total carbon and increasing pH. The comparison between geochemical modeling of $CO_2$ degassing and observed data showed that pH of observed ACW was higher than calculated pH. This can be happen when degassed $CO_2$ does not come out from the solution and/or exist in solution as $CO_{2(g)}$ bubble. In that case, $CO_{2(g)}$ bubble doesn't affect the pH and alkalinity. Thus alkalinity measured by ANT and NDIR-TC could not detect the $CO_2$ bubble although measured alkalinity was similar to the calculated alkalinity. Moreover, total carbon measured by ANT and NDIR-TC could be underestimated. Consequently, it is necessary to compare the alkalinity and total carbon data from various kind of methods and interpret very carefully. This study provide technical information of measurement of dissolve $CO_2$ from $CO_2$-rich water which could be natural analogue of geologic sequestration of $CO_2$.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.175-197
• /
• 2021

Recently, with the development of deep learning technology, research on unstructured data analysis is being actively conducted, and it is showing remarkable results in various fields such as classification, summary, and generation. Among various text analysis fields, text classification is the most widely used technology in academia and industry. Text classification includes binary class classification with one label among two classes, multi-class classification with one label among several classes, and multi-label classification with multiple labels among several classes. In particular, multi-label classification requires a different training method from binary class classification and multi-class classification because of the characteristic of having multiple labels. In addition, since the number of labels to be predicted increases as the number of labels and classes increases, there is a limitation in that performance improvement is difficult due to an increase in prediction difficulty. To overcome these limitations, (i) compressing the initially given high-dimensional label space into a low-dimensional latent label space, (ii) after performing training to predict the compressed label, (iii) restoring the predicted label to the high-dimensional original label space, research on label embedding is being actively conducted. Typical label embedding techniques include Principal Label Space Transformation (PLST), Multi-Label Classification via Boolean Matrix Decomposition (MLC-BMaD), and Bayesian Multi-Label Compressed Sensing (BML-CS). However, since these techniques consider only the linear relationship between labels or compress the labels by random transformation, it is difficult to understand the non-linear relationship between labels, so there is a limitation in that it is not possible to create a latent label space sufficiently containing the information of the original label. Recently, there have been increasing attempts to improve performance by applying deep learning technology to label embedding. Label embedding using an autoencoder, a deep learning model that is effective for data compression and restoration, is representative. However, the traditional autoencoder-based label embedding has a limitation in that a large amount of information loss occurs when compressing a high-dimensional label space having a myriad of classes into a low-dimensional latent label space. This can be found in the gradient loss problem that occurs in the backpropagation process of learning. To solve this problem, skip connection was devised, and by adding the input of the layer to the output to prevent gradient loss during backpropagation, efficient learning is possible even when the layer is deep. Skip connection is mainly used for image feature extraction in convolutional neural networks, but studies using skip connection in autoencoder or label embedding process are still lacking. Therefore, in this study, we propose an autoencoder-based label embedding methodology in which skip connections are added to each of the encoder and decoder to form a low-dimensional latent label space that reflects the information of the high-dimensional label space well. In addition, the proposed methodology was applied to actual paper keywords to derive the high-dimensional keyword label space and the low-dimensional latent label space. Using this, we conducted an experiment to predict the compressed keyword vector existing in the latent label space from the paper abstract and to evaluate the multi-label classification by restoring the predicted keyword vector back to the original label space. As a result, the accuracy, precision, recall, and F1 score used as performance indicators showed far superior performance in multi-label classification based on the proposed methodology compared to traditional multi-label classification methods. This can be seen that the low-dimensional latent label space derived through the proposed methodology well reflected the information of the high-dimensional label space, which ultimately led to the improvement of the performance of the multi-label classification itself. In addition, the utility of the proposed methodology was identified by comparing the performance of the proposed methodology according to the domain characteristics and the number of dimensions of the latent label space.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.18 no.4
• /
• pp.570-593
• /
• 1996

This study was intended to perform the influence of condyle positional change after surgical correction of skeletal Class III malocclusion after BSSRO in 20 patients(males 9, females 11) using computed tomogram that were taken in centric occlusion before, immediate, and long term after surgery and lateral cephalogram that were taken in centric occlusion before, 7 days within the period intermaxillary fixation, 24hour after removing intermaxillary fixation and long term after surgery. 1. Mean intercondylar distance was $84.45{\pm}4.01mm$ and horizontal long axis of condylar angle was $11.89{\pm}5.19^{\circ}$on right, $11.65{\pm}2.09^{\circ}$on left side and condylar lateral poles were located about 12mm and medial poles about 7mm from reference line(AA') on the axial tomograph. Mean intercondylar distance was $84.43{\pm}3.96mm$ and vertical axis angle of condylar angle was $78.72{\pm}3.43^{\circ}$on right, $78.09{\pm}6.12^{\circ}$on left. 2. No statistical significance was found on the condylar change(T2C-T1C) but it had definitive increasing tendency. There was significant decreasing of the distance between both condylar pole and the AA'(p<0.05) during the long term(TLC-T2C). 3. On the lateral cephalogram, no statistical significance was found between immediate after surgery and 24 hours after the removing of intermaxillary fixation but only the lower incisor tip moved forward about 0.33mm(p<0.05). Considering individual relapse rate, mean relapse rate was 1.2% on L1, 5.0% on B, 2.0% on Pog, 9.1% on Gn, 10.3% on Me(p<0.05). 4. There was statistical significance on the influence of the mandibular set-back to the total mandibular relapse(p<0.05). 5. There was no statistical significance on the influence of the mandibular set-back(T2-T1) to the condylar change(T2C-T1C), the condylar change(T2C-T1C, TLC-T2C) to the mandibular total relapse, the pre-operative condylar position to the condylar change(T2C-T1C, TLC-T2C), the pre-operative mandibular posture to the condylar change(T2C-T1C, TLC-T2C)(p>0.05). 6. The result of multiple regression analysis on the influence of the pre-operative condylar position to the total mandibular relapse revealed that the more increasing of intercondylar distance and condylar vertical axis angle and decreasing of condyalr head long axis angle, the more increasing of mandibular horizontal relapse(L1,B,Pog,Gn,Me) on the right side condyle. The same result was founded in the case of horizontal relapse(L1,Me) on the left side condyle.(p<0.05). 7. The result of multiple regression analysis on the influence of the pre-operative condylar position to the pre-operative mandibular posture revealed that the more increasing of intercondylar distance and condylar vertical axis angle and decreasing of condylar head long axis angle, the more increasing of mandibular vertical length on the right side condyle. and increasing of vertical lengh & prognathism on the left side condyle(p<0.05). 8. The result of simple regression analysis on the influence of the pre-operative mandibular posture to the mandibular total relapse revealed that the more increasing of prognathism, the more increasing of mandibular total relapse in B and the more increasing of over-jet the more increasing of mandibular total relapse(p<0.05). Consequently, surgical mandibular repositioning was not significantly influenced to the change of condylar position with condylar reposition method.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.135-149
• /
• 2020

Artificial intelligences are changing world. Financial market is also not an exception. Robo-Advisor is actively being developed, making up the weakness of traditional asset allocation methods and replacing the parts that are difficult for the traditional methods. It makes automated investment decisions with artificial intelligence algorithms and is used with various asset allocation models such as mean-variance model, Black-Litterman model and risk parity model. Risk parity model is a typical risk-based asset allocation model which is focused on the volatility of assets. It avoids investment risk structurally. So it has stability in the management of large size fund and it has been widely used in financial field. XGBoost model is a parallel tree-boosting method. It is an optimized gradient boosting model designed to be highly efficient and flexible. It not only makes billions of examples in limited memory environments but is also very fast to learn compared to traditional boosting methods. It is frequently used in various fields of data analysis and has a lot of advantages. So in this study, we propose a new asset allocation model that combines risk parity model and XGBoost machine learning model. This model uses XGBoost to predict the risk of assets and applies the predictive risk to the process of covariance estimation. There are estimated errors between the estimation period and the actual investment period because the optimized asset allocation model estimates the proportion of investments based on historical data. these estimated errors adversely affect the optimized portfolio performance. This study aims to improve the stability and portfolio performance of the model by predicting the volatility of the next investment period and reducing estimated errors of optimized asset allocation model. As a result, it narrows the gap between theory and practice and proposes a more advanced asset allocation model. In this study, we used the Korean stock market price data for a total of 17 years from 2003 to 2019 for the empirical test of the suggested model. The data sets are specifically composed of energy, finance, IT, industrial, material, telecommunication, utility, consumer, health care and staple sectors. We accumulated the value of prediction using moving-window method by 1,000 in-sample and 20 out-of-sample, so we produced a total of 154 rebalancing back-testing results. We analyzed portfolio performance in terms of cumulative rate of return and got a lot of sample data because of long period results. Comparing with traditional risk parity model, this experiment recorded improvements in both cumulative yield and reduction of estimated errors. The total cumulative return is 45.748%, about 5% higher than that of risk parity model and also the estimated errors are reduced in 9 out of 10 industry sectors. The reduction of estimated errors increases stability of the model and makes it easy to apply in practical investment. The results of the experiment showed improvement of portfolio performance by reducing the estimated errors of the optimized asset allocation model. Many financial models and asset allocation models are limited in practical investment because of the most fundamental question of whether the past characteristics of assets will continue into the future in the changing financial market. However, this study not only takes advantage of traditional asset allocation models, but also supplements the limitations of traditional methods and increases stability by predicting the risks of assets with the latest algorithm. There are various studies on parametric estimation methods to reduce the estimated errors in the portfolio optimization. We also suggested a new method to reduce estimated errors in optimized asset allocation model using machine learning. So this study is meaningful in that it proposes an advanced artificial intelligence asset allocation model for the fast-developing financial markets.

• Journal of Distribution Science
• /
• v.10 no.7
• /
• pp.33-41
• /
• 2012

The purpose of this study is to analyse the state of the energy industry and to determine the efficiency of its functioning on the basis of energy conservation principle and application of innovative technologies aimed at improving the ecological modernisation of agricultural sectors of Kazakhstan. The research methodology is based on an integrated approach of financial and economic evaluation of the effectiveness of the investment project, based on calculation of elasticity, total costs and profitability, as well as on comparative, graphical and system analysis. The current stage is characterised by widely spread restructuring processes of electric power industry in many countries through introduction of new technical installations of energy facilities and increased government regulation in order to enhance the competitive advantage of electricity market. Electric power industry features a considerable value of creating areas. For example, by providing scientific and technical progress, it crucially affects not only the development but also the territorial organisation of productive forces, first of all the industry. In modern life, more than 90% of electricity and heat is obtained by Kazakhstan's economy by consuming non-renewable energy resources: different types of coal, oil shale, oil, natural gas and peat. Therefore, it is significant to ensure energy security, as the country faces a rapid fall back to mono-gas structure of fuel and energy balance. However, energy resources in Kazakhstan are spread very unevenly. Its main supplies are concentrated in northern and central parts of the republic, and the majority of consumers of electrical power live in the southern and western areas of the country. However, energy plays an important role in the economy of industrial production and to a large extent determines the level of competitive advantage, which is a promising condition for implementation of energy-saving and environmentally friendly technologies. In these circumstances, issues of modernisation and reforms of this sector in Kazakhstan gain more and more importance, which can be seen in the example of economically sustainable solutions of a large local monopoly company, significant savings in capital investment and efficiency of implementation of an investment project. A major disadvantage of development of electricity distribution companies is the prevalence of very high moral and physical amortisation of equipment, reaching almost 70-80%, which significantly increases the operating costs. For example, while an investment of 12 billion tenge was planned in 2009 in this branch, in 2012 it is planned to invest more than 17 billion. Obviously, despite the absolute increase, the rate of investment is still quite low, as the total demand in this area is at least more than 250 billion tenge. In addition, industrial infrastructure, including the objects of Kazakhstan electric power industry, have a tangible adverse impact on the environment. Thus, since there is a large number of various power projects that are sources of electromagnetic radiation, the environment is deteriorated. Hence, there is a need to optimise the efficiency of the organisation and management of production activities of energy companies, to create and implement new technologies, to ensure safe production and provide solutions to various environmental aspects. These are key strategic factors to ensure success of the modern energy sector of Kazakhstan. The contribution of authors in developing the scope of this subject is explained by the fact that there was not enough research in the energy sector, especially in the view of ecological modernisation. This work differs from similar works in Kazakhstan in the way that the proposed method of investment project calculation takes into account the time factor, which compares the current and future value of profit from the implementation of innovative equipment that helps to bring it to actual practise. The feasibility of writing this article lies in the need of forming a public policy in the industrial sector, including optimising the structure of energy disbursing rate, which complies with the terms of future modernised development of the domestic energy sector.

• The Journal of Korean Society for Radiation Therapy
• /
• v.22 no.2
• /
• pp.135-144
• /
• 2010

Purpose: The purpose of this study is that through production of phantom for respiration gated radiotherapy, assessing appropriacy of exposure dose for the therapy using RPM (Real-time Position Management). Materials and Methods: We located measurement object on the phantom for respiration gated radiotherapy made of 2 linear actuator, acrylic panel, stanchion, iron plate ets. to drive (up, down, front, back). Using 4D CT scan, we analyzed patient's respiration and reproduced the movement by computer. On the phantom, we located a 2D-Array (PTW) and an White water phantom (4.5 cm) and used DMLC (interval 2 cm) in the field size $10{\times}10\;cm$, then exposed 21EX X-ray 100 MU, in the case of phantom was (1) static (2) moving (3) gated using RPM respectively gantry $0^{\circ}$ and $90^{\circ}$ We measured with a 0.125 CC ionization chamber (PTW) on the phantom (7.5 cm) in the same condition. Results: Ionization chamber: There were within 0.3% of error with gating respiration and approximately 2% of error without gating in the same condition. 2D-Array: Gantry $90^{\circ}$, field size $10{\times}10\;cm$, using DMLC. There were within 3% of error with gating respiration and approximately 16% of error without gating. Conclusion: The phantom for respiration gated radiotherapy makes plans considering patient's movement, quantitative analysis of exposure dose and proper assessment therapy for IMRT patients using RPM possible.

• The Korean Journal of Physiology
• /
• v.2 no.2
• /
• pp.101-135
• /
• 1968

As physical fitness measured was muscle strength (hand grips, leg extention, back lift, and arm pull and thrust), skinfold thickness (5 different sites), circulatory function (resting heart rate and blood pressure), speed (kinesiological analysis during 100m sprint, record, maximal and final speed), motor function (50 meter dash, ball throwing, standing broad jump, and pull-ups), maximal aerobic power (maximum oxygen intake by field running method), muscle power (leg and arm by inertia ergometer), and general endurance (maximum endurance running time on the treadmill at the speed of 5 MPH and grade of 15.5%) of 1131 Korean children (boys 572, girls 559) aged of 6 to 17 years, who were randomly sampled from 24 primary, middle and high schools at the two districts of Seoul and KyungKi. The results are summarized as follows: 1) The status (height and weight) of the children was almost same as that of the previously reported Korean and Japanese children of same ages. 2) Muscle strength was a gained linearly with geting age in the boys and girls but there was a little improvement in girls aged of 13 years or more. 3) The mean skin fold thickness was increased linearly with geting ages in both sexes, but the girls from 12 to 17 years of age were increased rapidly, and maximum value was 17mm, while boys was 7.0 mm. 4) In the circulatory function, the resting heart rate was decreased, but the blood pressure was increased with ages in both sexes within the normal limits. 5) The maximum and final speed during 100 meter sprint increases with age in boys but girls who are 12 years old or older, were not improved any mere. The patterns of running were same in both sexes, and maximum speed reached at about 30 meters from starting line. 6) The motor function was increased with age in both sexes, but there was no improvement in 12 years of age or older girls. More over records of all functions except standing broad jump was less than those of Japanese in the same age, respectively. 7) The maximum oxygen intake (MOI) was increased considerably with ages and maximum values were 2.93 L/min (boys) and 2.09 L/min (girls) at the age of 17years. This result was almost same as that of the Japanese and Easter Island population, but the value was lower than that of Europe. The average of the maximum oxygen intake per kg body weight per minute from 9 to 17 years of age were around 53 ml in the boys and 42 ml in the girls. 8) Muscle power was increased linearly with ages in boys while there was relatively a little increment in girls. The maximum values of leg muscle in boys and girls at the 17 years of age were 0.168 and 0.088 horse power, respectively. 9) The maximum endurance running time was increased considerably from the age of 9 in boys, while there was no improvement in girls. The maximum values were 6.0 min and 1.8 min, respectively.