• Title/Summary/Keyword: hyper method

Search Result 386, Processing Time 0.034 seconds

Fault Detection in LDPE Process using Machine Learning Techniques (머신러닝 기법을 활용한 LDPE 공정의 이상 감지)

  • Lee, Changsong;Lee, Kyu-Hwang;Lee, Hokyung
    • Korean Chemical Engineering Research
    • /
    • v.58 no.2
    • /
    • pp.224-229
    • /
    • 2020
  • We propose a machine learning-based method for proactively detecting faults in LDPE processes and predicting equipment lifespan. It is important to detect and prevent unexpected faults in chemical processes in order to maximize safety and productivity. Since LDPE process is a high-pressure process up to 3,000 kg/㎠g or more, once ESD occurs, it can result in productivity loss due to increased maintenance periods. By collecting key variables operation data of the process and using unsupervised machine leaning methods, we developed a fault detection model which detected 4 ESDs 2.4 days prior to the occurrence. In addition, it was confirmed that the life expectancy of a hyper compressor can be predicted by using the physically significant key variables.

(Searching Effective Network Parameters to Construct Convolutional Neural Networks for Object Detection) (물체 검출 컨벌루션 신경망 설계를 위한 효과적인 네트워크 파라미터 추출)

  • Kim, Nuri;Lee, Donghoon;Oh, Songhwai
    • Journal of KIISE
    • /
    • v.44 no.7
    • /
    • pp.668-673
    • /
    • 2017
  • Deep neural networks have shown remarkable performance in various fields of pattern recognition such as voice recognition, image recognition and object detection. However, underlying mechanisms of the network have not been fully revealed. In this paper, we focused on empirical analysis of the network parameters. The Faster R-CNN(region-based convolutional neural network) was used as a baseline network of our work and three important parameters were analyzed: the dropout ratio which prevents the overfitting of the neural network, the size of the anchor boxes and the activation function. We also compared the performance of dropout and batch normalization. The network performed favorably when the dropout ratio was 0.3 and the size of the anchor box had not shown notable relation to the performance of the network. The result showed that batch normalization can't entirely substitute the dropout method. The used leaky ReLU(rectified linear unit) with a negative domain slope of 0.02 showed comparably good performance.

Improved Purification Process for Cholera Toxin and its Application to the Quantification of Residual Toxin in Cholera Vaccines

  • Jang, Hyun;Kim, Hyo-Seung;Kim, Jeong-Ah;Seo, Jin-Ho;Carbis, Rodney
    • Journal of Microbiology and Biotechnology
    • /
    • v.19 no.1
    • /
    • pp.108-112
    • /
    • 2009
  • A simplified method for the purification of cholera toxin was developed. The 569B strain of Vibrio cholerae, a recognized hyper-producer of cholera toxin, was propagated in a bioreactor under conditions that promote the production of the toxin. The toxin was separated from the bacterial cells using 0.2-${\mu}m$ crossflow microfiltration, the clarified toxin was passed through the membrane into the permeate, and the bacterial cells were retained in the retentate. The 0.2-${\mu}m$ permeate was then concentrated 3-fold and diafiltered against 10 mM phosphate buffer, pH 7.6, using 30-kDa crossflow ultrafiltration. The concentrated toxin was loaded onto a cation exchange column, the toxin was bound to the column, and most of the impurities were passed unimpeded through the column. The toxin was eluted with a salt gradient of phosphate buffer, pH 7.0, containing 1.0 M NaCl. The peak containing the toxin was assayed for cholera toxin and protein and the purity was determined to be 92%. The toxin peak had a low endotoxin level of $3.1\;EU/{\mu}g$ of toxin. The purified toxin was used to prepare antiserum against whole toxin, which was used in a $G_{M1}$ ganglioside-binding ELISA to determine residual levels of toxin in an oral inactivated whole-cell cholera vaccine. The $G_{M1}$ ganglioside-binding ELISA was shown to be very sensitive and capable of detecting as little as 1 ng/ml of cholera toxin.

An Analysis about the Behavior of the Wiper Blade Including Incompressibility (비압축성을 고려한 와이퍼 블레이드의 거동 해석)

  • Chung, Won-Sun;Song, Hyun-Seok;Park, Tae-Won;Jung, Sung-Pil;Kim, Wook-Hyeon
    • Transactions of the Korean Society of Automotive Engineers
    • /
    • v.18 no.2
    • /
    • pp.83-90
    • /
    • 2010
  • The windshield wiper consists of 4 parts: a blade, an arm, a linkage and a motor. The wiper blade makes contact with the windshield and is designed to be operated normally at an angle of 30~50 degrees to the front glass. If the contact pressure between the wiper blade and windshield surface is too high, noise and wear of the rubber will result. On the other hand, if the contact pressure is too low, the performance will do badly, since foreign substances such as dust and stains will not be removed well. The pressure and friction of the wiper blade has a great influence on its effectiveness in cleaning the front window. This is due to the contact of the rubber with the window. This paper presents the dynamic analysis method to estimate the performance of the flat type blade of the wiper system. The blade has a nonlinear characteristic since the rubber is an incompressible hyper-elastic and visco-elastic material. Thus, Structural dynamic analysis using a complex contact model for the blade is performed to find the characteristics of the blade. The flexible multi-body dynamic model is verified by the comparison between test and analysis result. Also, the optimization using the central composite design table is performed.

Direct Brow Lift Combined with Suspension of the Orbicularis Oculi Muscle

  • Lee, Jeong Woo;Cho, Byung Chae;Lee, Kyung Young
    • Archives of Plastic Surgery
    • /
    • v.40 no.5
    • /
    • pp.603-609
    • /
    • 2013
  • Background Although the conventional direct brow lift operation provides a simple means of managing lateral brow ptosis, the scars produced have been unacceptable. However, using the modifications proposed here, scarring showed remarkable improvement. This article reviews our experiences with the presented technique, mainly with respect to postoperative scarring. Methods Measured amounts of supra-eyebrow skin and subcutaneous fat were excised en bloc in the conventional manner under 'hyper-hydrated' local infiltration anesthesia. The lower flap and the edge of the upper flap were undermined above the muscular plane, and the orbicularis oculi muscle was directly suture-plicated and suspended upward to the distal frontalis muscle. Skin closure was performed in a basic plastic surgical manner. Results From April 2007 to April 2012, a consecutive series of 60 patients underwent surgery using the above method. The average width of the excised skin was 8 mm (range, 5-15 mm) at the apex of the eyebrow. Preoperative complaints were resolved without occurrence of significant complications. The surgical scars showed remarkable improvement and were negligible in the majority of the cases. Conclusions The direct brow lift operation combined with plication/suspension of the superior and lateral portion of the orbicularis oculi muscle provides a simple, safe, and predictable means of correcting lateral brow ptosis. The scars were acceptable to all of the patients. For proper management of the frontalis tone, upper blepharoplasty and/or repair of eyelid levator function must be considered in addition to brow lift procedures.

Can Non-aqueous Solvent Desalinate?: Suggestion of the Screening Protocol for Selection of Potential Solvents (비수용성 용매를 이용한 탈염화 가능한가?: 적용 가능한 용매선정 기법 제안)

  • Choi, Oh Kyung;Seo, Jun Ho;Kim, Gyeong Soo;Kim, Dooil;Lee, Jae Woo
    • Journal of Korean Society on Water Environment
    • /
    • v.36 no.1
    • /
    • pp.48-54
    • /
    • 2020
  • This paper presents a screening protocol for the selection of solvents available for the solvent extraction desalination process. The desalination solvents hypothetically and theoretically require the capability of (1) Forming hydrogen bonds with water, (2) Absorbing some water molecules into its non-polar solvent layer, (3) Changing solubility for water-solvent separation, and (4) Rejecting salt ions during absorption. Similar to carboxylic acids, amine solvents are solvent chemicals applicable for desalination. The key parameter for selecting the potential solvent was the octanol-water partitioning coefficient (Kow) of which preferable value for desalination was in the range of 1-3. Six of the 30 amine solvents can absorb water and have a variable, i.e., temperature swing solubility with water molecule for water-solvent separation. Also, the hydrogen bonding interaction between solvent and water must be stronger than the ion-dipole interaction between water and salt, which means that the salt ions must be broken from the water and only water molecules absorbed for the desalination. In the final step, three solvents were selected as desalination solvents to remove salt ions and recover water. The water recovery of these three solvents were 15.4 %, 2.8 %, 10.5 %, and salt rejection were 76 %, 98 %, 95 %, respectively. This study suggests a new screening protocol comprising the theoretical and experimental approaches for the selection of solvents for the desalination method which is a new and challenges the desalination process in the future.

Improving Weak Classifiers by Using Discriminant Function in Selecting Threshold Values (판별 함수를 이용한 문턱치 선정에 의한 약분류기 개선)

  • Shyam, Adhikari;Yoo, Hyeon-Joong;Kim, Hyong-Suk
    • The Journal of the Korea Contents Association
    • /
    • v.10 no.12
    • /
    • pp.84-90
    • /
    • 2010
  • In this paper, we propose a quadratic discriminant analysis based approach for improving the discriminating strength of weak classifiers based on simple Haar-like features that were used in the Viola-Jones object detection framework. Viola and Jones built a strong classifier using a boosted ensemble of weak classifiers. However, their single threshold (or decision boundary) based weak classifier is sub-optimal and too weak for efficient discrimination between object class and background. A quadratic discriminant analysis based approach is presented which leads to hyper-quadric boundary between the object class and background class, thus realizing multiple thresholds based weak classifiers. Experiments carried out for car detection using 1000 positive and 3000 negative images for training, and 500 positive and 500 negative images for testing show that our method yields higher classification performance with fewer classifiers than single threshold based weak classifiers.

A Visual Expression in Fashion Illustration using 2D Graphics (2D 그래픽스를 활용한 패션 일러스트레이션의 시각적 표현 양상)

  • Choi Jung-Hwa
    • The Research Journal of the Costume Culture
    • /
    • v.13 no.4 s.57
    • /
    • pp.550-563
    • /
    • 2005
  • These days, photoshop and illustrator program can make a fashion illustration express easily and speedily, And they can also express a feeling and sensitivity of fashion illustrator by a tool and effect more than a manual work's media. The purpose of this study was to analyze a visual expression and characteristics in fashion illustration using 2D graphics. The method of this study was to analyze the fashion illustration books using 2D graphics. The visual expressions in fashion illustration using 2D graphics were as follows: Fashion illustration was based on a sketch or photography, and used a composition, drawing, mapping, painting, and manual work's re-touching. Characteristics of visual expression were as follows: First, a image composition was showed discontinuity and heterogeneity of image, new context and composition, and allowance of reality. Second, image transform was showed image overlap, body transformation by image recomposition, and deformed transformation by vector drawing. Third, hyper-real was showed precise touching, mechanical and neutral image, omission of background and focus of an object's characteristic and information. Fourth, following a realistic expression was showed simplified color, shading, dominant view point of fashion illustrator by omission, and daily lift style. Fifth, following a pictorial expression was showed non-fixed and irregular line, natural painting, and drawing and painting by conventional pictorial media. In conclusion, a photoshop and illustrator in 2D graphics will serve as a new media far fashion illustration with a manual work. And they will not only intensify a capacity as a commercial role of fashion illustration but also present a positive motive for students learning a fashion illustration.

  • PDF

Historical Study of Glamour Style (글래머 스타일의 사적 고찰)

  • Park, Ju-Hee
    • The Research Journal of the Costume Culture
    • /
    • v.16 no.2
    • /
    • pp.382-396
    • /
    • 2008
  • The purpose of this study is to analyze the transition of glamour style expressed in modern fashion throughout the history. The glamour style in modern fashion was categorized into five periods which are Hollywood glamour(H), Feminine glamour(F), Rich glamour(R), Decadent glamour(D), and Trash glamour(T). These categories were analyzed and compared in the viewpoint of aesthetic values deduced on the previous study. As a method of analysis, literature study and case study through the publications in aesthetics, history and preceding theses were used. The results of analysis are: Ostentatious luxury was suggested by extravagance of the movie star in H, the splendid American lifestyle in F, conspicuous consumption in R, fin de siecle bad taste in D and self-assured exposure in T. Mysterious idolatry was studied as exaggerated goddesshood of the movie star in H. It was expanded to celebrities including fashion models, television actresses, pop singers and young couture designers in the other periods. Artificial sensuality was originated from the femme fatale image of the courtesan as well as the traditional femininity of Victorian era in H. It was developed through exaggerated hourglass silhouette of the wife dressing in F. But an aggressiveness in R, a hyper-sexual femme fatale in D, an independent and defiant image of the showgirl in T were observed. Playful queerness stood out clearly by the fin de siecle phenomena. Though it was embodied in sexual perversive subculture, it emerged as hi fashion by young designers in the late R and became a crucial aesthetic value throughout D and T. It is more connected with the origins of glamour which arouses sexual ambiguity, crudity, aggressiveness and death. The continuity and discontinuity of the glamour style based on the theory of 'linked solution' were also analyzed.

  • PDF

Adaptive Customer Relation Management Strategies using Association Rules (연관 규칙을 이용한 적응적 고객 관계 관리 전략)

  • Han, Ki-Tae;Chung, Kyung-Yong;Baek, Jun-Ho;Kim, Jong-Hun;Ryu, Joong-Kyung;Lee, Jung-Hyun
    • Proceedings of the Korea Contents Association Conference
    • /
    • 2008.05a
    • /
    • pp.84-86
    • /
    • 2008
  • The customer relation marketing in which companies can utilize to control and to get the filtered information efficiently has appeared. It is applying data mining to build the management that can even predict and recommend products to customers. In this paper, we proposed the adaptive customer relation management strategies using the association rules of data mining. The proposed method uses the association rules composes frequent customers with occurrence of candidate customer set creates the rules of associative customers. We analyzed the efficient feature of purchase customers using the hyper graph partition according to the lift of creative association rules. Therefore, we discovered strategies of the cross-selling and the up-selling about customers.

  • PDF