Papers by Keyword: Combination Method

Abstract: Boosting algorithms are a means of building a strong ensemble classifier by aggregating a sequence of weak hypotheses. An ensemble consists of a set of independently trained classifiers whose predictions are combined when classifying novel instances. Previous research has shown that an ensemble as a whole is often more accurate than any of the single classifiers in the ensemble. In this paper we use applied technology to built an ensemble using a voting methodology of Boosting-BAN and Boosting-MultiTAN ensembles with 10 sub-classifiers in each one. We performed a comparison with Boosting-BAN and Boosting-MultiTAN ensembles with 25 sub-classifiers on standard benchmark datasets and the proposed technique was the most accurate. These results argue that boosting algorithms deserve more attention in machine learning and data mining communities.

Abstract: This paper proposes a more accurate springback prediction method of ageing forming for 2124 aluminum alloy. In age forming of panels, pre-bending radius, aging time and wall thickness of panels are selected as three parameters, make use of uniform design to arrange experiment and obtain springback radius using ABAQUS simulation. By means of regression analysis, the data is processed to get the influence caused by parameters on springback radius. Regression and BP neural network forecasting method are used respectively to predict springback radius and maximum prediction error is less than 31%. Combination method based on BP neural network is adopted and this method gets the satisfying prediction results that prediction error is within 5%. So conclusion can be drawn that prediction accuracy of combination method is much better than that of regression and BP neural network forecasting.

Abstract: An ensemble consists of a set of independently trained classifiers whose predictions are combined when classifying novel instances. Previous research has shown that an ensemble as a whole is often more accurate than any of the single classifiers in the ensemble. Boosting-BAN classifier is considered stronger than Boosting-MultiTAN on noise-free data. However, there are strong empirical indications that Boosting-MultiTAN is much more robust than Boosting-BAN in noisy settings. For this reason, in this paper we built an ensemble using a voting methodology of Boosting-BAN and Boosting-MultiTAN ensembles with 10 sub-classifiers in each one. We performed a comparison with Boosting-BAN and Boosting-MultiTAN ensembles with 25 sub-classifiers on standard benchmark datasets and the proposed technique was the most accurate.

Abstract: In order to deal with the difficulty of magnetic field adjustment in permanent magnet synchronous generators (PMSG), this paper proposed a novel paratactic structure hybrid excitation synchronous generators (HESG) and presented the basic configuration and its principle. Then, it analyzed three combination methodsof PMSG and the electro-excitation synchronous generators (EESG) in HESG,and also presented calculation way of field current in HESG. Finally, the experimental results proved the correctness of those analyses.

Abstract: In order to analyze inter-harmonic parameters accurately, a new detecting method was proposed on the basis of nonlinear optimization and AR modern spectral estimation. The AR model was used to produce a rough estimated value of harmonic parameters and then the signal model was set. By using the optimization algorithm which based on damped nonlinear least squares method and conjugate gradient method, the parameters can be estimated more accurately. The combined method not only solves the problem of inaccurate estimate by AR power spectrum analysis, but also solves the problem of how to set the signal model and to choose initial value. The simulation results show that in noisy environments, this method is effective and the accuracy which is 101 to 104 times higher than the interpolation Hanning window method.

Abstract: The residual fatigue life of a submarine pressure structure is investigated, based on the combination between the methods of conventional Monte Carlo and classical probabilistic fracture mechanics. Firstly, Monte Carlo method is employed to obtain the reliability of given initial fatigue life. Secondly, the two induced factors M_A1 and M_A2 in the paper are estimated according to the initial fatigue life and the reliability. Thirdly, based on the two factors, the residual fatigue life based on other reliability is obtained by using classical probabilistic fracture mechanics method. Numerical examples show that the proposed method is more efficient without accuracy loss for residual fatigue life compared with Monte Carlo method. This method can also be employed to predict the residual fatigue life on other analogue structures.

Abstract: A new method is presented in the paper. The fatigue life reliability of submarine cone-cylinder shell is investigated, based on the combination between the methods of conventional Monte Carlo and classical probabilistic fracture mechanics. Firstly, Monte Carlo method is employed to obtain the reliability of given initial fatigue life. Secondly, the two induced factors M1 and M2 in the paper are estimated according to the initial fatigue life and the reliability. Thirdly, based on the two factors, the other fatigue life reliability is obtained by using classical probabilistic fracture mechanics method. Finally, numerical cases show that the proposed method is more efficient without accuracy loss for fatigue life reliability compared with Monte Carlo method. This method can also be applied to predict the fatigue life reliability of analogue structures.

Abstract: All the construction methods of subway stations are firstly reviewed in the world, which are cut and cover method, mine tunneling method, tunnel boring machine method and combination methods of them. Finally characteristics of various construction methods are analyzed in detail.

Abstract: The combination method implied in the formula used to evaluate the building separation to avoid seismic pounding in current Taiwan Building Code (TBC) is the absolute sum (ABS) method. However, it was demonstrated that the ABS method is not satisfactory. In this study, the seismic pounding risk of buildings is used as an indicator to demonstrate the validity of application of various combination methods to the formula in the pounding related provisions of TBC. The result shows that the building separations based on these combination methods generally provide relatively conservative estimate due to an overestimate property line setback.

Abstract: A novel combination method is proposed in order to efficiently combine conflict evidence. The credibility of the evidence is derived by introducing a distance function, and the conflict probability is distributed to every proposition according to its credibility. Better performance is demonstrated by numerical examples compared with three other methods.