Search results

The Factor/PCA node provides powerful data-reduction techniques to reduce the complexity of our data.
Example of Data Mining Examining the Data.
Data Preparation.
Data Mining with SQL Server 2008.
Data Mining with SQL Server 2005.

According to conditions of the irrigation area water supply and groundwater reduction for many years, the paper puts forward the optimal allocation and the supplying water programs on using combination of surface water and groundwater.
Taking the 2001 - 2010 average annual groundwater in agricultural production uses 6.912 billion m3 is regarded as the initial value, design four kinds of groundwater reduction scenarios, reduced by 3%, 5%, 10%, 15% of agricultural groundwater extraction, agriculture needs Yellow River water respectively reaching 4.20736, 4.3456, 4.6912, 5.0368 billion m3 / a[3].
However, according to the data in recent years, 2006 - 2010 Yellow River Basin water has decreased, about three percent less than the same period average value.
Changing some parameters value in the model (each parameter is selected based on the experimental data or empirical data), it could work out the predetermined needs diverting water from the Yellow River and groundwater extraction at different time.
Calculating data will have better reference value that will increase the new water-saving irrigation area, improve the efficiency of irrigation water resource utilization, the rational distribution exploitation quantity of groundwater, effectively control the expansion of shallow groundwater funnel area, solve groundwater recharge and balance, analyze carrying capacity of water resources and so on.

With only 3 classifiers mining traffic data from 3 different aspects, the regularities can be fully digged out.
Process of feature selection The traffic data usually has high dimensions and redundant features.
So the data need to do dimension reduction which means the operation of feature selection.FCBF (Fast Correlation-Based Filter Solution) by reference [5] is a fast and effective method to produce feature subset.
It calculates the feature used to divide the data set to make data purity declines the most.
Experiments and Analysis Data Resource Experimental data is from Moore et.al which contains 377 526 network flows divided into 10 types; each flow consists of 249 features, of which the last feature means the category.

data[12].
Symbols:experimental data[6].
J Chem Eng Data, 1992, 37: 96-100
Data, 2010, 55 (11), 4580-4585
Data, 2008, 53(11), 2521-2531.

The temperature increase to 700°C, has been transformed a part of the orthorhombic ZrO2 to a monoclinic phase, contributing to a surface area reduction of the samples, showing irregular agglomerates in morphology, with adsorption/desorption isotherms type IV and mesoporosity characteristic.
It can be seen on Fig. 1 that main peaks of tetragonal and orthorhombic phases are very close to each other, with a difference of just dq = 0,09o between their angles, according to the crystallographic chips JCPDF 79-1769 JCPDF 79-1769 (tetragonal zirconia) and JCPDF 79-1796 (orthorhombic zirconia) from data package of Shimadzu program.
The data of X-rays diffraction collected were used for phase identification and crystallite size calculation from the extension line of X-ray (d111) through the deconvolution of secondary diffraction line of polycrystalline cerium (used as standard), using the Scherrer equation [12].
For these calculations, it was used the major phase density identified in the XRD and it was considered theoretical density (ρ) of 5.606 g/cm³ for monoclinic ZrO2 and 6.047 g/cm³ for orthorhombic ZrO2, obtained according to the crystallographic chips JCPDF 65-2357 and JCPDF 79-1796 from data package of Shimadzu program.
It was also found that all samples have shown peaks displacement to the orthorhombic and monoclinic phase to smaller angles, when compared with the crystallographic chips JCPDF 65-2357 and JCPDF 79-1796 from data packet of Shimadzu program and the displacement value was less evident in the sample calcined at 700°C.

Data Acquisition and Pre-processing To automatically generate the trajectory of spray gun, data acquisition is a primary step.
Its pre-processing process includes removal of abnormal data, data interpolation, data smoothing and data reduction.
However, limited by the density of point cloud data, it is impossible to get enough data points on the plane to generate a complete contour.
After the same operation on each slice, a set of slice data is obtained.
The spraying position is obtained by the average sampling of slice data.

For this purpose, KPCA [12] was conducted to the input data.
Afterwards the training data were projected onto these selected components.
For a simplest linear 2-class issue, if some given training data points each belong to one of two classes, the machine can separate the two class data by a hyperplane which has the largest margin to the nearest data point of each class.
In that space, the maximum-margin hyperplane can be constructed easily to discriminate the mapped data points.
Data mining and knowledge discovery, 1998, 2(2): 121-167

SVM supports small sample of data, but also has advantages as low computational complexity, high sparse model, support for multiple classification and class membership probability calculation and so on.
As shown in Fig.1, we choose high fidelity bullet pinhole type noise reduction micro sound listener to collect acoustic signal in circuit breaker operation.
Fig.3 Acoustic signal EEMD decomposition envelope 3 Experiment Analysis Data of this paper comes from the collection of data LW59-252 type SF6 circuit breaker, normal, institutional jam, lack of lubrication of the experiment data of acoustic data under three kinds of breaker status. 3.1 Feature Extraction According to the fault feature extraction method mentioned in section 2, calculate energy entropy IMF under various states as shown in Table 1.
Taking experimental data of high voltage circuit breaker were calculated energy entropy of EEMD decomposition as feature vector input support vector machine, half of the data which is used as the training data, the other half as the test data set which is used to test the correctness of diagnosis.
At the same time, the SVM method can be applied to the fault diagnosis of high voltage circuit breaker which is small sample data set, and construct more reliable model through choosing different kernel function. 4 Conclusion In this paper, the high voltage circuit breaker fault diagnosis based on acoustic data is mainly studied because of there is very strong sound signals when high voltage circuit breaker acts, the acoustic signal will change while the circuit breaker motion state changes.

In the current study, grain-scale variant selection is described in terms of quantitative data analysis.
Instead of using conventional ODF's, here Rodrigues-Frank space is employed to represent the orientation data.
If each data point in the orientation map is considered as a "grain" with a specific orientation, the data can provide information about the "texture" of the scanned area.
The intensities were calculated from 1.3 million data points, of which about 96% were from the bcc and 4% from the fcc phases.
Here, a recent dislocation-based model [3, 9] is evaluated using the orientation data obtained from individual γ grains.

In order to maintain better approximation, it adopts testing data of typical stress path, testing data of uniaxial tension and torsion test.
Today in many fields such as aeronautics, astronautics, marine, transportation, machinery, construction engineering, disaster prevention, disaster reduction and accident prevention, heat stress and heat effect have been put more attention.