• International Journal of Computer Science & Network Security
• /
• 제23권8호
• /
• pp.190-198
• /
• 2023

To enhance customer satisfaction for higher profits, an e-commerce sector can establish a continuous relationship and acquire new customers. Utilize machine-learning models to analyse their customer's behavioural evidence to produce their competitive advantage to the e-commerce platform by helping to improve overall satisfaction. These models will forecast customers who will churn and churn causes. Forecasts are used to build unique business strategies and services offers. This work is intended to develop a machine-learning model that can accurately forecast retainable customers of the entire e-commerce customer data. Developing predictive models classifying different imbalanced data effectively is a major challenge in collected data and machine learning algorithms. Build a machine learning model for solving class imbalance and forecast customers. The satisfaction accuracy is used for this research as evaluation metrics. This paper aims to enable to evaluate the use of different machine learning models utilized to forecast satisfaction. For this research paper are selected three analytical methods come from various classifications of learning. Classifier Selection, the efficiency of various classifiers like Random Forest, Logistic Regression, SVM, and Gradient Boosting Algorithm. Models have been used for a dataset of 8000 records of e-commerce websites and apps. Results indicate the best accuracy in determining satisfaction class with both gradient-boosting algorithm classifications. The results showed maximum accuracy compared to other algorithms, including Gradient Boosting Algorithm, Support Vector Machine Algorithm, Random Forest Algorithm, and logistic regression Algorithm. The best model developed for this paper to forecast satisfaction customers and accuracy achieve 88 %.

• 한국콘크리트학회:학술대회논문집
• /
• 한국콘크리트학회 2003년도 가을 학술발표회 논문집
• /
• pp.617-620
• /
• 2003

This paper presents the confinement effect of transverse reinforcement and cross-tie in hollow rectangular sectional columns. 20 analytical models with different amounts of transverse reinforcement and cross-tie in a plastic hinge region were analyzed by 3D nonlinear FEM. The analytical results show that the higher ductility can be achieved by the resonable combination of transverse re-bar and cross-tie providing sufficient lateral confinement.

• 한국콘크리트학회:학술대회논문집
• /
• 한국콘크리트학회 1998년도 가을 학술발표대회 논문집(III)
• /
• pp.821-826
• /
• 1998

The design methods for the steel plate strengthened RC structures are not well established yet because the structural behavior of plated RC beams is more complex than that of regular unplated ones. The main purpose of this paper is to present the premature failure mechanism of steel plate strengthened RC beams. The analytical models of interfacial stress and normal are also proposed in this paper. The comparisons between the analytical results using the proposed theory and experimental ones relatively satisfied.

• Earthquakes and Structures
• /
• 제7권2호
• /
• pp.201-216
• /
• 2014

In this paper the vulnerability of the confined masonry buildings is evaluated analytically. The proposed approach includes the nonlinear dynamic analysis of the two-story confined masonry buildings with common plan as a reference structure. In this approach the damage level is calculated based on the probability of exceedance of loss vs a specified ground motion in the form of fragility curves. The fragility curves of confined masonry wall buildings are presented in two levels of limit states corresponding to elastic and maximum strength versus PGA based on analytical method. In this regard the randomness of parameters indicating the characteristics of the building structure as well as ground motion is considered as likely uncertainties. In order to develop the analytical fragility curves the proposed analytical models of confined masonry walls in a previous investigation of the authors, are used to specify the damage indices and responses of the structure. In order to obtain damage indices a series of pushover analyses are performed, and to identify the seismic demand a series of nonlinear dynamic analysis are conducted. Finally by considering various mechanical and geometric parameters of masonry walls and numerous accelerograms, the fragility curves with assuming a log normal distribution of data are derived based on capacity and demand of building structures in a probabilistic approach.

• 한국안전학회지
• /
• 제32권1호
• /
• pp.66-74
• /
• 2017

Masonry walls which are commonly used for partitions in low-rise reinforced concrete (RC) structures, can be easily exposed to high risks under strong earthquakes. Since the strength degradations cannot be protected under the ground motions, their applications cannot be recommended for building structures which are designed to possess high seismic performances. However, masonry-infilled walls are typically considered as non-structural elements in evaluating the seismic performance of building structures. In order to figure out this problem, this study performed experiments using two specimens-only RC frame and RC frame infilled with masonry walls- under static loading. Also, the study established analytical models representing fully infilled frames and bare frame, and compared their structural behavior with test results. In addition, analytical model representing partially infilled frames was established and analyzed. Test results indicated that strength and energy dissipating capacity were increased for IW-RN(fully infilled frames) compared to the NW(bare frame). The nonlinear static analysis of the three specimens was also conducted using the inelastic plastic hinge frame element and diagonal strut models, and the analytical results successfully simulated the nonlinear behaviour of the specimens in accordance with the test results.

• Structural Engineering and Mechanics
• /
• 제65권5호
• /
• pp.535-546
• /
• 2018

The objective of this study is to understand the behaviour of hollow core slabs strengthened with FRP and hybrid techniques through numerical and analytical studies. Different strengthening techniques considered in this study are (i) External Bonding (EB) of Carbon Fiber Reinforced Polymer (CFRP) laminates, (ii) Near Surface Mounting (NSM) of CFRP laminates, (iii) Bonded Overlay (BO) using concrete layer, and (iv) hybrid strengthening which is a combination of bonded overlay and NSM or EB. In the numerical studies, three-dimensional Finite Element (FE) models of hollow core slabs were developed considering material and geometrical nonlinearities, and a phased nonlinear analysis was carried out. The analytical calculations were carried out using Response-2000 program which is based on Modified Compression Field Theory (MCFT). Both the numerical and analytical models predicted the behaviour in agreement with experimental results. Parametric studies indicated that increase in the bonded overlay thickness increases the peak load capacity without reducing the displacement ductility. The increase in FRP strengthening ratio increased the capacity but reduced the displacement ductility. The hybrid strengthening technique was found to increase the capacity of the hollow core slabs by more than 100% without compromise in ductility when compared to their individual strengthening schemes.

• Steel and Composite Structures
• /
• 제26권6호
• /
• pp.759-769
• /
• 2018

This paper studies shear and tensile behaviors of headed stud connectors in double skin composite (DSC) structure. Firstly, 11 push-out tests and 11 tensile tests were performed to investigate the ultimate shear and tensile behaviors of headed stud in DSC shear wall, respectively. The main parameters investigated in this test program were height and layout of headed stud connectors. The test results reported the representative failure modes of headed studs in DSC structures subjected to shear and tension. The shear-slip and tension-elongation behaviors of headed studs in DSC structures were also reported. Influences of different parameters on these shear-slip and tension-elongation behaviors of headed studs were discussed and analyzed. Analytical models were also developed to predict the ultimate shear and tensile resistances of headed stud connectors in DSC shear walls. The developed analytical model incorporated the influence of the dense layout of headed studs in DSC shear walls. The validations of analytical predictions against 22 test results confirmed the accuracy of developed analytical models.

• Water Engineering Research
• /
• 제2권3호
• /
• pp.187-196
• /
• 2001

An approximate analytical solution of a nonlinear hydro-thermo coupled diffusion equation is derived using the dimensionless form of the equation and transformation method. To derive an analytical solution, it is drastically assumed that the product of first order derivatives in the non-dimensionalized governing equation has little influence on the solution of heat and moisture behavior problem. The validity of this drastic assumption is demonstrated. Some numerical simulation is performed to investigate the applicability of a derived approximate analytical solution. The results show a good agreement between analytical and numerical solutions. The proposed solution may provide a useful tool in the verification process of the numerical models. Also, the solution can be used for the analysis of one-dimensional coupled heat and moisture movements in unsaturated porous media.

• 한국산학기술학회논문지
• /
• 제17권9호
• /
• pp.724-731
• /
• 2016

지진으로 인한 구조물의 피해가 지속적으로 증가하면서, 구조물의 취약성을 평가하는 일은 지진 대비에 필수적으로 여겨지고 있다. 지진 취약도 곡선은 지진에 대한 구조물의 안전도에 대한 확률 지표로써 널리 이용되고 있으며, 많은 연구자들에 의해 보다 정확하고 효율적인 취약도 곡선 도출을 위한 노력이 계속되고 있다. 하지만 기존의 대부분의 연구에서는 취약도 곡선 도출시 수치해석 시간 절약을 위해 단순화된 2차원 해석모델을 사용해 왔는데, 많은 경우에 있어 2차원 모델은 정확한 구조물의 내진 거동 및 지진 취약성을 평가하기에 적당하지 않을 수 있다. 이에 본 연구에서는 3차원 해석 모델을 사용하여 더욱 정확하면서도 여전히 효과적으로 지진 취약도 곡선을 도출할 수 있는 방법을 제시한다. 이 방법은 신뢰성 해석 소프트웨어인 FERUM과 구조해석 소프트웨어인 ZEUS-NL을 서로 연동시켜 상호 자동적인 데이터 교환이 가능하게 하고, 샘플링 기법이 아닌 FORM 해석 기법을 통해 구조물의 파괴확률을 구한다. 이는 3차원 모델을 사용의 경우에도 효율적으로 구조 신뢰성 해석이 가능하게 해준다. 이를 이용해 RC 프레임 구조물의 3차원 해석 모델을 사용하여 지진 취약성 평가를 수행하였다.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제11권4호
• /
• pp.262-271
• /
• 2011

In this work, we present a unified analytical surface potential model, valid for both PD and FD SOI MOSFETs. Our model is based on a simplified one dimensional and purely analytical approach, and builds upon an existing model, proposed by Yu et al. [4], which is one of the most recent compact analytical surface potential models for SOI MOSFETs available in the literature, to improve its accuracy and remove its inconsistencies, thereby adding to its robustness. The model given by Yu et al. [4] fails entirely in modeling the variation of the front surface potential with respect to the changes in the substrate voltage, which has been corrected in our modified model. Also, [4] produces self-inconsistent results due to misinterpretation of the operating mode of an SOI device. The source of this error has been traced in our work and a criterion has been postulated so as to avoid any such error in future. Additionally, a completely new expression relating the front and back surface potentials of an FD SOI film has been proposed in our model, which unlike other models in the literature, takes into account for the first time in analytical one dimensional modeling of SOI MOSFETs, the contribution of the increasing inversion charge concentration in the silicon film, with increasing gate voltage, in the strong inversion region. With this refinement, the maximum percent error of our model in the prediction of the back surface potential of the SOI film amounts to only 3.8% as compared to an error of about 10% produced by the model of Yu et al. [4], both with respect to MEDICI simulation results.