• Korean Journal of Child Studies
• /
• v.1
• /
• pp.65-81
• /
• 1980

This study is aimed at investigating the compensatory education which was already implemented or is being implemented in the U.S.A. and Japan; and at studying the types of programs and their characteristics; and at sounding out the possibilities of the application of such programs in family and social conditions is Korea. In order to achieve the above mentioned objectives, the established items for the study are as follows: (1) Various types of early children's education (2) Programs of compensatory education for the disadvantaged Children (3) Head Start Program, Early Training Project and Montessori School (4) Integrated Preschool Programs (5) Day-Care Center for employed mothers We investigated the various compensatory education programs for the preschool children who are in economically, socially, culturally disadvantaged conditions. Head Start Programs were federally supported programs for preschool children and opened as summer programs in 1965 for the first time. The purpose of Head Start has been to give preschool children the kinds of experiences they need in preparation for school. The Head Start children were found to be significantly better prepared for school than the normal children. However, after six to eight months, their initial advantages had virtually. disappeared and then the simple problem with Head Start and other such programs was that little long-term good could be evidenced unless the high quality educational environment was maintained. Therefore, to solve this problem, three other programs were funded as part of the overall Head Start. These three programs are the Parent-Child Center, Home Start, and the Child and Family Resources Program. The Early Training Project for disadvantaged children was implemented by Klaus and Gray of Peabody College in 1962. The program was a field research study concerned with the development and testing over time of procedures for improving the educability of young children from low income homes. Its major concern was to study whether it was possible to offset the progressive retardation observed in the public schooling careers of children, living in deprived circumstances. Children, who were trained through the Early Training Project were superior to control groups in the test of IQ and vocabulary as well as linguistic abilities, and preparation for reading. This project showed the possibilities which could prevent preschool children from being disadvantaged socially, culturally and mentally. In 1907, Montessori School was established by Maria Montessori in Italy and her school program has been introduced at present to several countries in the world as one compensatory educations. She first began her experimental methods with retarded children, followed by disadvantaged children from the tenements of Rome. The Montessori approach futures a prepared environment and carefully designed, self-correcting materials. The Montessori curriculum presents tastes that feature sequence, order, and regularity, in addition to those that develop motor and sensory skills. She was interested in children's intellectual development and in developing good work habits. One of the latest developed programs for disadvantaged children is "Integrated Preschool Program" which has successfully integrated handicapped and nonhandicapped children. Several studies have showed that handicapped children in integrated school environments are accepted by and interact with their nonhandicapped peers. In fact, this program provides a number of potential, and perhaps opportunities for nonhandicapped children to serve as valuable resources in fostering the development of their handicapped peers. Next we turn to Japanese programs which are divided into two different types. One is Day-Care Center which was established by Child Welfare Law and the other is kindergarten organized by School Education Law. The kindergarten opened in 1876 and it has been part of school systems since 1947 by the implementation of education law, and the Day-Care Center which started in 1890 for the employed mothers. was changed into Day-Nursery by the enactment of child welfare law in 1947. The laws and operational regulations for the Day-Nursery were set up and were put in effect by the establishment standard acts of children welfare facilities, and the Day-Nursery has been operated in various types by the increasing demand, chiefly because of the socio-economical changes of family structures in both urban and suburban areas. Nursery education for physically and mentally disadvantaged children is for those who are blind, deaf and dumb, mentally retarded; physically disadvantaged by accidents or diseases. Montessori education in Japan was started in 1968 and many research groups for studying Montessori were organized. In 1977, Montessori remedial education society was also organized in which they started a number of studies; a study for developing materials; in-service training for the remedial education; and seminars and lectures, etc It is strongly suggested that we study the early educations that are being implemented in Japan and a variety of compensatory educations that were already implemented in the U.S.A. and modify them for the organization of our own model and properly accommodate them to our social needs.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2008.06a
• /
• pp.239-240
• /
• 2008

As semiconductor devices are scaled down for better performance and more functionality, the Cu-based interconnects suffer from the increase of the resistivity of the Cu wires. The resistivity increase, which is attributed to the electron scattering from grain boundaries and interfaces, needs to be addressed in order to further scale down semiconductor devices [1]. The increase in the resistivity of the interconnect can be alleviated by increasing the grain size of electroplating (EP)-Cu or by modifying the Cu surface [1]. Another possible solution is to maximize the portion of the EP-Cu volume in the vias or damascene structures with the conformal diffusion barrier and seed layer by optimizing their deposition processes during Cu interconnect fabrication, which are currently ionized physical vapor deposition (IPVD)-based Ta/TaN bilayer and IPVD-Cu, respectively. The use of in-situ etching, during IPVD of the barrier or the seed layer, has been effective in enlarging the trench volume where the Cu is filled, resulting in improved reliability and performance of the Cu-based interconnect. However, the application of IPVD technology is expected to be limited eventually because of poor sidewall step coverage and the narrow top part of the damascene structures. Recently, Ru has been suggested as a diffusion barrier that is compatible with the direct plating of Cu [2-3]. A single-layer diffusion barrier for the direct plating of Cu is desirable to optimize the resistance of the Cu interconnects because it eliminates the Cu-seed layer. However, previous studies have shown that the Ru by itself is not a suitable diffusion barrier for Cu metallization [4-6]. Thus, the diffusion barrier performance of the Ru film should be improved in order for it to be successfully incorporated as a seed layer/barrier layer for the direct plating of Cu. The improvement of its barrier performance, by modifying the Ru microstructure from columnar to amorphous (by incorporating the N into Ru during PVD), has been previously reported [7]. Another approach for improving the barrier performance of the Ru film is to use Ru as a just seed layer and combine it with superior materials to function as a diffusion barrier against the Cu. A RulTaN bilayer prepared by PVD has recently been suggested as a seed layer/diffusion barrier for Cu. This bilayer was stable between the Cu and Si after annealing at $700^{\circ}C$ for I min [8]. Although these reports dealt with the possible applications of Ru for Cu metallization, cases where the Ru film was prepared by atomic layer deposition (ALD) have not been identified. These are important because of ALD's excellent conformality. In this study, a bilayer diffusion barrier of Ru/TaCN prepared by ALD was investigated. As the addition of the third element into the transition metal nitride disrupts the crystal lattice and leads to the formation of a stable ternary amorphous material, as indicated by Nicolet [9], ALD-TaCN is expected to improve the diffusion barrier performance of the ALD-Ru against Cu. Ru was deposited by a sequential supply of bis(ethylcyclopentadienyl)ruthenium [Ru$(EtCp)_2$] and $NH_3$plasma and TaCN by a sequential supply of $(NEt_2)_3Ta=Nbu^t$ (tert-butylimido-trisdiethylamido-tantalum, TBTDET) and $H_2$ plasma. Sheet resistance measurements, X-ray diffractometry (XRD), and Auger electron spectroscopy (AES) analysis showed that the bilayer diffusion barriers of ALD-Ru (12 nm)/ALD-TaCN (2 nm) and ALD-Ru (4nm)/ALD-TaCN (2 nm) prevented the Cu diffusion up to annealing temperatures of 600 and $550^{\circ}C$ for 30 min, respectively. This is found to be due to the excellent diffusion barrier performance of the ALD-TaCN film against the Cu, due to it having an amorphous structure. A 5-nm-thick ALD-TaCN film was even stable up to annealing at $650^{\circ}C$ between Cu and Si. Transmission electron microscopy (TEM) investigation combined with energy dispersive spectroscopy (EDS) analysis revealed that the ALD-Ru/ALD-TaCN diffusion barrier failed by the Cu diffusion through the bilayer into the Si substrate. This is due to the ALD-TaCN interlayer preventing the interfacial reaction between the Ru and Si.

• Korean Journal of Heritage: History & Science
• /
• v.47 no.4
• /
• pp.74-91
• /
• 2014

The stone statues in the site of Mireuksaji Temple(Iksan, South Korea) were created after the stone pagoda was built in 639. They, transitional statues between animal sculptures and human shaped statues made from the late Goryeo dynasty until the early Joseon dynasty, were set up at the four corners of the stone pagoda by way of guardians. In the case of three statues, their surfaces were denudated and their iconographies have been indiscernible. However, the one in the southwest clearly shows its iconography. It is inferior in properties to the other three statues in the northwest, the northeast and the southeast respectively, but on the other hand its iconography has been well maintained. The reason is related to exposure to harmful environments; specifically, the retaining wall, built around the stone pagoda in the 17th century, had the southwest statue inside and could naturally worked as a buffer against harmful environments. As a result, for about 400 years there has been difference in weathering conditions between the three stone statues and the southwest statue, which brought denudation, the consequent indiscernibleness of iconography and biological invasion to the three statues, notwithstanding superior properties(northwest statue:$176kgf/cm^2$, northeast statue:$109kgf/cm^2$, southeast statue:$273kgf/cm^2$). In contrast, the southwest statue significantly shows its iconography with black contaminants and granule decomposition, despite inferior properties($133kgf/cm^2$). Defenseless exposure to external environment is not recommended for the stone statues, because it is hard to preserve the extant iconography. Herein lies the application of the data on microclimate around Mireuksaji Temple. As regards the weathering zone in which the stone statues are located, Conservation increases in acidity and frequency as years go by, Hereat, in the approach to the Conservation of stone statues, the first consideration needs to be morphological historicity rather than geographical location.

• Journal of Wetlands Research
• /
• v.25 no.2
• /
• pp.145-158
• /
• 2023

Constructed wetlands (CWs) are effective technologies for urban wastewater management, utilizing natural physico-chemical and biological processes to remove pollutants. This study employed a bibliometric analysis approach to investigate the progress and future research trends in the field of CWs. A comprehensive review of 100 most-recently published and open-access articles was performed to analyze the performance of CWs in treating wastewater. Spain, China, Italy, and the United States were among the most productive countries in terms of the number of published papers. The most frequently used keywords in publications include water quality (n=19), phytoremediation (n=13), stormwater (n=11), and phosphorus (n=11), suggesting that the efficiency of CWs in improving water quality and removal of nutrients were widely investigated. Among the different types of CWs reviewed, hybrid CWs exhibited the highest removal efficiencies for BOD (88.67%) and TSS (95.67%), whereas VSSF, and HSSF systems also showed high TSS removal efficiencies (83.25%, and 78.83% respectively). VSSF wetland displayed the highest COD removal efficiency (71.82%). Generally, physical processes (e.g., sedimentation, filtration, adsorption) and biological mechanisms (i.e., biodegradation) contributed to the high removal efficiency of TSS, BOD, and COD in CW systems. The hybrid CW system demonstrated highest TN removal efficiency (60.78%) by integrating multiple treatment processes, including aerobic and anaerobic conditions, various vegetation types, and different media configurations, which enhanced microbial activity and allowed for comprehensive nitrogen compound removal. The FWS system showed the highest TP removal efficiency (54.50%) due to combined process of settling sediment-bound phosphorus and plant uptake. Phragmites, Cyperus, Iris, and Typha were commonly used in CWs due to their superior phytoremediation capabilities. The study emphasized the potential of CWs as sustainable alternatives for wastewater management, particularly in urban areas.

• Journal of Ethics
• /
• no.77
• /
• pp.73-112
• /
• 2010

This study aims to explore the character and problem of multicultural society in Korea in terms of the concept of moral panics. Its major issues are dealt with as follows: Firstly, this essay will apply two concepts of culture in multiculturalism - the pattern of meaning and a study of perfection- to three degrees of multicultural members: by individual, by groups and by a society as a whole. In this approach, moral panics of multicultural society in Korea have been manipulated by the secondary definitions like Korean government and media. In this study, however, the resource of the panics would be seen as nationalism in Korean history. To remove it in this essay, the conception of the pattern of meaning, which makes members understand others outer their norms, should be harmonized with that of a study of perfection by which they have identities. Secondly, the main subject of multiculturalism in Korea should at least be majority (groups)-Korean, not minority (groups)-foreigners. A stereotype of foreigners by majority is an image distorted by nationalities and races. People, for example, with the white skin from advanced countries are recognized as superior, while those born in the countries of Southeast Asia are, consciously or unconsciously, discriminated and have low positions due to socio-economic stratification in Korea. In this sense, a study of multicultural society in Korea should go forward to the inner direction to majority, because it is one of the real moral panics in Korea. In conclusion, it is important that there must be a study of identity which we can have of others in multicultural studies of Korea. It enables us to meet the conception of diversity. In that Korean government and media have neglected the danger of nationalism, it is also necessary that this study have any foundation of morality in ethics, which can give useful alternatives to the given polices of the secondary definitions.

• Asia Marketing Journal
• /
• v.13 no.3
• /
• pp.183-208
• /
• 2011

After more than 25 years of accumulated research evidence, there is little doubt that leadership behavior is related to a wide variety of positive individual and organizational outcomes. Indeed, leadership behavior has been empirically linked to increased employee satisfaction, organizational commitment, extra effort, turnover intention, organizational citizenship behavior, and overall employee performance. However, it is important to point out that although leadership behavior has been linked to a number of positive organizational outcomes, research regarding the antecedents of such behavior is limited. Especially there is little research dealing with the antecedents of inter-organizational leadership behavior. Supply chain leadership can be defined as the activities undertaken by the supply chain leader to influence the management programs and strategies of supply chain members. Supply chain performance is influenced by leadership of supply chain leader. Although research on supply chain leadership can be broadly categorized, many researchers have been preoccupied with analyzing supply chain leadership by the power-influence approach measuring such as control, power, and power bases. Also they have not examined the relationship between leadership and financial performance. This study has started to overcome those research gaps. The purpose of this study is to investigate the effect of relational behaviors on supply chain leadership, and the effect of such leadership behavior on financial performance of supply chain followers. In addition, this study also try to find out moderating variable existing in the relationship. To be concrete, First, this study develops a model of the antecedents of four conceptually distinct forms of relational behaviors such as training, fair reward, offering vision, and inter-organizational communication, and tests the hypothesized differential effects of relational behavior forms on supply chain leadership. Second, this study tests the effect of supply chain leadership on financial performance. Third, this study investigates the extent to which this leadership-performance relationship is moderated by leader ethicality. The reason why this study deals with convenience store supply chain is that there is very strong inter-dependence between a franchisor and its suppliers. Their strong inter-dependence makes their relationship as the relationship between a superior and subordinates and creates an atmosphere that leadership occur without difficulty. For the purpose of empirical testing, 217 respondents of suppliers of convenience store supply chain in Korea were surveyed and the analysis utilizing partial least square model indicated that training, fair reward, inter-organizational communication had positive effects on supply chain leadership, and such leadership had positive effect on financial performance of followers. On the other hand, offering vision had no effect on supply chain leadership. In addition, leader ethicality had moderating effect on the relationship between supply chain leadership and financial performance.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.18 no.5
• /
• pp.91-107
• /
• 2023

Lean startup is a concept that combines the words "lean," meaning an efficient way of running a business, and "startup," meaning a new business. It is often cited as a strategy for minimizing failure in early-stage businesses, especially in software-based startups. By scrutinizing the case of a startup L, this study suggests that lean startup methodology(LSM) can be useful for hardware and manufacturing companies and identifies ways for early startups to successfully implement LSM. To this end, the study explained the core of LSM including the concepts of hypothesis-driven approach, BML feedback loop, minimum viable product(MVP), and pivot. Five criteria to evaluate the successful implementation of LSM were derived from the core concepts and applied to evaluate the case of startup L . The early startup L pivoted its main business model from defecation alert system for patients with limited mobility to one for infants or toddlers, and finally to a smart bottle for infants. In developing the former two products, analyzed from LSM's perspective, company L neither established a specific customer value proposition for its startup idea and nor verified it through MVP experiment, thus failed to create a BML feedback loop. However, through two rounds of pivots, startup L discovered new target customers and customer needs, and was able to establish a successful business model by repeatedly experimenting with MVPs with minimal effort and time. In other words, Company L's case shows that it is essential to go through the customer-market validation stage at the beginning of the business, and that it should be done through an MVP method that does not waste the startup's time and resources. It also shows that it is necessary to abandon and pivot a product or service that customers do not want, even if it is technically superior and functionally complete. Lastly, the study proves that the lean startup methodology is not limited to the software industry, but can also be applied to technology-based hardware industry. The findings of this study can be used as guidelines and methodologies for early-stage companies to minimize failures and to accelerate the process of establishing a business model, scaling up, and going global.

• International conference on construction engineering and project management
• /
• 2024.07a
• /
• pp.1286-1288
• /
• 2024

The construction industry, known for its inherent risks and multiple hazards, necessitates effective solutions for hazard identification and mitigation [1]. To address this need, the implementation of machine learning models specializing in object detection has become increasingly important because this technological approach plays a crucial role in augmenting worker safety by proactively recognizing potential dangers on construction sites [2], [3]. However, the challenge in training these models lies in obtaining accurately labeled datasets, as conventional methods require labor-intensive labeling or costly measurements [4]. To circumvent these challenges, synthetic data generation (SDG) has emerged as a key method for creating realistic and diverse training scenarios [5], [6]. The paper reviews the evolution of synthetic data generation tools, highlighting the shift from earlier solutions like Synthpop and Data Synthesizer to advanced game engines[7]. Among the various gaming platforms, Unity 3D and Unreal Engine stand out due to their advanced capabilities in replicating realistic construction hazard environments [8], [9]. Comparing Unity 3D and Unreal Engine is crucial for evaluating their effectiveness in SDG, aiding developers in selecting the appropriate platform for their needs. For this purpose, this paper conducts a comparative analysis of both engines assessing their ability to create high-fidelity interactive environments. To thoroughly evaluate the suitability of these engines for generating synthetic data in construction site simulations, the focus relies on graphical realism, developer-friendliness, and user interaction capabilities. This evaluation considers these key aspects as they are essential for replicating realistic construction sites, ensuring both high visual fidelity and ease of use for developers. Firstly, graphical realism is crucial for training ML models to recognize the nuanced nature of construction environments. In this aspect, Unreal Engine stands out with its superior graphics quality compared to Unity 3D which typically considered to have less graphical prowess [10]. Secondly, developer-friendliness is vital for those generating synthetic data. Research indicates that Unity 3D is praised for its user-friendly interface and the use of C# scripting, which is widely used in educational settings, making it a popular choice for those new to game development or synthetic data generation. Whereas Unreal Engine, while offering powerful capabilities in terms of realistic graphics, is often viewed as more complex due to its use of C++ scripting and the blueprint system. While the blueprint system is a visual scripting tool that does not require traditional coding, it can be intricate and may present a steeper learning curve, especially for those without prior experience in game development [11]. Lastly, regarding user interaction capabilities, Unity 3D is known for its intuitive interface and versatility, particularly in VR/AR development for various skill levels. In contrast, Unreal Engine, with its advanced graphics and blueprint scripting, is better suited for creating high-end, immersive experiences [12]. Based on current insights, this comparative analysis underscores the user-friendly interface and adaptability of Unity 3D, featuring a built-in perception package that facilitates automatic labeling for SDG [13]. This functionality enhances accessibility and simplifies the SDG process for users. Conversely, Unreal Engine is distinguished by its advanced graphics and realistic rendering capabilities. It offers plugins like EasySynth (which does not provide automatic labeling) and NDDS for SDG [14], [15]. The development complexity associated with Unreal Engine presents challenges for novice users, whereas the more approachable platform of Unity 3D is advantageous for beginners. This research provides an in-depth review of the latest advancements in SDG, shedding light on potential future research and development directions. The study concludes that the integration of such game engines in ML model training markedly enhances hazard recognition and decision-making skills among construction professionals, thereby significantly advancing data acquisition for machine learning in construction safety monitoring.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Radiation Oncology Journal
• /
• v.16 no.3
• /
• pp.265-274
• /
• 1998

Purpose : This prospective study has been conducted to assess the value of three dimensional conformal radiation therapy (3DCRT) for lung cancer and to determine its potential advantage over current treatment approaches. Specific aims of this study were to 1) find the most ideal 3DCRT technique 2) establish the maximum tolerance dose that can be delivered with 3DCRT and 3) identify patients at risk for development of radiation pneumonitis. Materials and Methods : Beginning in Nov. 1994, 95 patients with inoperable non-small cell lung cancer (stage I; 4, stage II; 1, stage IIIa; 14, stage IIIb; 76) were entered onto this 3D conformal trial Areas of known disease and elective nodal areas were initially treated to 45 Gy and then using 3DCRT technique 65 to 70 Gy of total dose were delivered to the gross disease. Sixty nine patients received 65 Gy of total dose and 26 received 70 Gy Seventy eight patients (82.1$\%$) also received concurrent MVP chemotherapy. 3DCRT plans were compared with 2D plans to assess the adequacy of dose delivery to target volume, dose volume histograms for normal tissue, and normal tissue complication Probabilities (NTCP). Results : Most of plans (78/95) were composed of non-coplanar multiple (4-8) fields. Coplanar segmented conformal therapy was used in 17 pateints, choosing the proper gantry angle which minimize normal lung exposure in each segment. 3DCRT gave the full dose to nearly 100$\%$ of the gross disease target volume in all patients. The mean NTCP for ipsilateral lung with 3DCRT (range; 0.17-0.43) was 68$\%$ of the mean NTCP with 2D treatment planning (range; 0.27-0.66). DVH analysis for heart showed that irradiated volume of heart could be significantly reduced by non-coplanar 3D approach especially in the case of left lower lobe lesion. Of 95 patients evaluable for response, 75 (79$\%$), showed major response including 25 (26$\%$) with complete responses and 50 (53$\%$) with partial responses. One and two rear overall survivals of stage III patients were 62.6$\%$ and 35.2$\%$ respectively. Twenty percent (19/95) of patients had pneumonitis; Eight patients had grade 1 pneumonitis and 11 other patients had grade 2. Comparison of the average of NTCP for lung showed a significant difference between patients with and without radiation pneumonitis. Average NTCP for Patients without complication was 62$\%$ of those with complications. Conclusions : This study showed that non-coplanar multiple fields (4-8) may be one of the ideal plans for 3DCRT for lung cancer. It also suggested that 3DCRT may provide superior delivery of high dose radiation with reduced risk to normal tissue and that NTCP can be used as a guideline for the dose escalation.