Search results

The selection of training set need data under normal operation, preventing data-affected training results.
Step 3 Dispose the data of training set, and code data according to unified standard.
Split data set into several split unit.
Step 8 Reduce function operates SVM data.
Meanwhile, extract the same parameters of equipment system and put data into actual data set.

To reduce the complexity of caculation, it is necessary to extract the feature of sample data before the modeling of neural network, that is, in terms of a certain evaluation criteria, the best feature subset is selected from input feature set.
Suppose the independent variables are , the dependent variable is , and take sample points to construct the data table and .
Then can represent data table as good as possible, and has the best explanation ability to the dependent variable .
Table 1 show the partial information of selected data sets.
Table 1 Partial information of selected data set Table 2 Modleing ability comparison of three data set Conclusion PLS can reduce the input dimensions, noise pollution and multiple correlations between independent variables, namely it has a better feature extraction ability.

The aim of this study is to acquire more quantitative data on the cold-deformed state, in order to determine the key parameters for nucleation (like orientation gradient, dislocation density, degree of fragmentation…) and to correlate these to parameters which can be predicted by a mean field deformation model.
The degree of banding seems unaffected by rolling temperature in IF material and affects one third of the grains after 65% reduction, as seen by Barnett [5].
The hot band with an average grain size of 20µm had been deformed by cold-rolling with different thickness reductions in a laboratory rolling mill.
The first stages will be studied here, corresponding to 15%, 29%, 42% and 51% in thickness reduction.
Fig. 4: Proportion of grains belonging to αfibre, γ-fibre (±15°) and grains of all other orientations depending on the section reduction.

On - line monitoring and state reduction The focus of the overall concept is the on-line state space reduction allowing dynamical system monitoring.
Fig. 3 The state space reduction on the example of a tank In the second step of the concept, on-line state space reduction was carried out.
In the third and last step, the reduction was evaluated. 3.
It represents the loading and processing analyser data.
Within the on-line state space reduction, qualitative model components are reduced based on dynamic models and on data obtained from sensors and actuators.

Predicted results are compared with available experimental data.
There is an interesting feature in stiffness degradation approach that only limited amount of data is needed for obtaining reasonable results [11]. [12] reports that the reduction of bending strength of foam cored sandwich specimen is caused by the stiffness reduction of foam due to ageing of polyurethane foam during fatigue cycles. [13] investigate the static and flexural fatigue characteristics of foam core polymer composite sandwich beams.
Predicted results are compared to the available experimental data.
If such a relationship could be established, extrapolation of data at various stress levels would be derived, thus the amount of test data needed to characterize the materials behaviour [16].
Fatigue data were generated at load levels of 100, 90, 80, 70 and 60% of the static ultimate load.

The percentage reduction of the bacteria reached >99.9%.
The percentage reduction of bacterial colonies was taken as a measure of the antibacterial activity.
Surface elemental analysis data from EDS elements before sputtering after sputtering Weight percentage Atom percentage Weight percentage Atom percentage C 47.67 54.86 45.69 53.64 O 52.13 45.04 52.14 45.96 Si 0.20 0.10 0.31 0.15 Ag 0.00 0.00 1.86 0.24 (a) substrate before sputtering (b) substrate after sputtering at 180 W power Fig.2 EDS spectra of the samples.
Table 2 summarises the measured antibacterial activity of samples prepared at powers from 120 to 200 W: the percentage reduction of Staphylococcus aureus bacteria reached >99.9%.
Summary Regenerated cellulose films coated with nano silver by RF magnetron sputtering possessed antibacterial activity: the percentage reduction of Staphylococcus aureus bacteria in the presence of silver coated film reached >99.9%.

How to play the advantage of gradient data, improve the detection effect, is a subject for gradient data processing and interpretation need studying.
Magnetic data are often acquired on an undulating surface.
Most existing techniques for data enhancement and interpretation require data on a horizontal plane, the reduction of observed magnetic data values on an arbitrary surface to a level plane is necessary before quantitative analysis.
Analytic signal and Euler deconvolution method are also require data on a horizontal plane.
Reid et al. (1990) developed the 3D Euler deconvolution, operating on grided magnetic data[5].

The XRD data show the changes of the structure and phase composition of titanium dioxide thin films due to the nitrogen doping.
The reduction of crystallites size takes place at the increase of the nitrogen concentration.
For the phase analysis of the thin films the PDF-4 database of International Center for Diffraction Data (ICDD) was used.
The data calculated from XRD measurement for nitrogen-doped TiO2 films are shown in Table 2.
The crystallites size reduction with the increasing of the nitrogen content agreed with our previous calculations with cross-sectional method from SEM data [17].

In order to obtain bulk nanoceramics, data of the form in Fig. 2 suggest that it is necessary to enhance the kinetics of densification and retard the kinetics of grain growth.
Unfortunately, there is not much quantitative data available for triple point diffusion.
Figure 4b shows experimental data from careful experiments on triple point drag in tricrystals; the data show that triple junction mobility has a higher activation energy than grain boundary mobility [10].
Recently, the following diffusion data have been reported for cation diffusion in fine grained 3YTZ [16]
Although diffusion creep mechanisms lead to a stress exponent of 1, data on 3YTZ with grain sizes of ~0.3 to 0.5 µm typically show a stress dependence of ~2 [17,18].

The main problem is the multiplicity of parameter estimation solutions arising from using spatially sparse and noised-polluted data.
Almost all feature extraction procedures perform some form of data reduction.
The measured data is compressed into small dimension feature vectors.
The main problem is the multiplicity of parameter estimation solutions arising from using spatially sparse and noised-polluted data.
The main problem is the multiplicity of parameter estimation solutions arising from using spatially sparse and noised-polluted data.