Fisher discriminant analysis (FDA) is a popular method for supervised dimensionality reduction. FDA seeks for an embedding transformation such that the ratio of the between-class scatter to the within-class scatter is...
详细信息
Fisher discriminant analysis (FDA) is a popular method for supervised dimensionality reduction. FDA seeks for an embedding transformation such that the ratio of the between-class scatter to the within-class scatter is maximized. Labeled data, however, often consume much time and are expensive to obtain, as they require the efforts of human annotators. In order to cope with the problem of effectively combining unlabeled data with labeled data to find the embedding transformation, we propose a novel method, called subspace semi-supervised Fisher discriminant analysis (SSFDA), for semi-supervised dimensionality reduction. SSFDA aims to find an embedding transformation that respects the discriminant structure inferred from the labeled data and the intrinsic geometrical structure inferred from both the labeled and unlabeled data. We also show that SSFDA can be extended to nonlinear dimensionality reduction scenarios by applying the kernel trick. The experimental results on face recognition demonstrate the effectiveness of our proposed algorithm.
TiO2 film is an important photo-induced catalyzer and it also presents hyper hydrophilicity properties. However the applications of this material are greatly limited due to its low efficiency under visible light. A pr...
详细信息
TiO2 film is an important photo-induced catalyzer and it also presents hyper hydrophilicity properties. However the applications of this material are greatly limited due to its low efficiency under visible light. A promising way to improve its photocatalytic effect is to dope with some nonmetal elements. The visible light absorption and hydrophilicity of N-doped TiO2 film on quartz substrates prepared by R.F magnetron sputtering are obtained, with the working gas of Ar/N2, the samples were post-treated by annealing under different temperatures in nitrogen atmosphere. The structure, doping, surface morphology and optical properties of undoped and N-doped TiO2 films by means of XRD, XPS, SEM and UV-vis spectra were analyzed. The XRD results indicated that TiO2 anatase is formed around 600°C and N atoms are doped in TiO2 crystal lattice as shown in XPS studies. For the UV-vis analysis of N-doped samples, it is found that a dramatic red drift occur in the absorbance spectra compared with that of TiO2 thin films, which means a higher visible light catalysis activity. The water contact angle measurement of the annealed samples under visible light radiation for 30 min showed excellent hydorphilicity with a water contact angle less than 10°.
Least squares support vector machine (LS-SVM) combined with niche genetic algorithm (NGA) are proposed for nonlinear sensor dynamic modeling. Compared with neural networks, the LS-SVM can overcome the shortcomings of ...
详细信息
Least squares support vector machine (LS-SVM) combined with niche genetic algorithm (NGA) are proposed for nonlinear sensor dynamic modeling. Compared with neural networks, the LS-SVM can overcome the shortcomings of local minima and over fitting, and has higher generalization performance. The sharing function based niche genetic algorithm is used to select the LS-SVM parameters automatically. The effectiveness and reliability of this method are demonstrated in two examples. The results show that this approach can escape from the blindness of man-made choice of LS-SVM parameters. It is still effective even if the sensor dynamic model is highly nonlinear.
This paper evaluates the validated uncertainty in SEVA sensor by integrating fault detection, identification and reconstruction (FDIR) and reliability engineering. The impact of each fault mode on measurement quality ...
详细信息
This paper evaluates the validated uncertainty in SEVA sensor by integrating fault detection, identification and reconstruction (FDIR) and reliability engineering. The impact of each fault mode on measurement quality is evaluated quantitatively by using a priori sensor reliability information to investigate the impact of incomplete fault coverage, FDIR and manual maintenance intervention. Bayesian probabilistic approach and uncertainty calculus are employed to model the impact of sensor validation on parameter uncertainty and to fuse the individual modes into a complete sensor model. A simulation of SEVA pressure sensor example illustrates the concept and conclusions.
In order to solve the precision guidance problem in television guidance missile system which is caused by time difference between the video frame used by the console and the video frame used by the guided missile. Thi...
详细信息
Model Predictive control (MPC) is one of the most powerful methods in modern control theory. In contrast to standard controllers, the model of the plant to be controlled is not only used for the design of the controll...
详细信息
A new prototype of self-validating pressure sensor is described. The elastic body of the transducer is based on circular-flat diaphragm structure. Eight groups of strain gauges are distributed on the diaphragm. One gr...
详细信息
In this paper we introduce a networked communication architecture for the design and development of pc-based robot control systems. Because of their performance and flexibility Pc-based solutions are more and more dep...
详细信息
ISBN:
(纸本)9781615677146
In this paper we introduce a networked communication architecture for the design and development of pc-based robot control systems. Because of their performance and flexibility Pc-based solutions are more and more deployed in the robotic control. On the other side the rising complexity of the today and future control tasks (i.e integration of singularity prediction modules or adaptronic component to compensate the vibration of the manipulator by high-speed handling) require considerable computing power, due to the time expensive data processing algorithms. By a predetermined control performance, this will lead to bottleneck when the required computing power exceeds the available CPU resource. The proposed networked is an extension (through distribution) of the current architecture deployed in the scope of the Collaborative research center SFB562 at the university of Braunschweig. It aims to extend the available computing power while keeping the hard real time constraints of the control tasks, even in the distributed environment.
this paper proposed a novel controlling technique of pulse width modulation (PWM) mode and pulse frequency modulation (PFM) mode to keep the high efficiency within width range of loading. The novel control method is u...
详细信息
A new current sensing circuit with bulk control (CCB) technique is presented in this paper. Compared to several current sensing techniques, there is a trade-off between the sensing accuracy and speed. Thus, the propos...
详细信息
暂无评论