Search results

The ID of a data set is commonly defined as the smallest number of dimensions that can be used to adequately explain the data.
In order to evaluate our proposed method in a real-world situation, seven data sets are selected from the UC Irvine machine learning data repository [13] with numeric features and varying data sizes, including several of high-dimensional data sets.
A summary of data sets can be found in Table 2.
This can be attributed to information loss owing to dimension reduction and the weak correlation structure in ’Glass’.
Muoh, “Data discretization unification,” the Seventh IEEE International Conference on Data Mining (ICDM Best Paper), pp. 183–192, 2007

Fig.4 The change of two tensile test yield Fig.5 The change of two tensile test tensile strength strength Fig.6 The change of two tensile test elongation Fig.7 The change of two tensile test yield ratio The yield strength and tensile strength of two kinds of steel yield are almost the same under the conditions of almost the same heating time and pressure reduction contrasting 1 with 3.Contrasting the both with 4,we find that yield strength and tensile strength will be increased significantly when the rolling process is reduced, 6 and 7 can also come to the conclusion.Comparing with the same reduction for 4 and 7,we find that the longer the heating time is, the higher the yield strength and tensile strength will be, at the same time, 1, 3 and 6 can also prove this point.
And elongation and tensile strength are concerned, tensile ability is strong, the deformation produce difficultly, low elongation, tensile ability is weak, the deformation produce easily, elongation is on the high side,this data can be reflected.
were cut to 25mm×15mm×10mm cube specimens with abrasive wheel cutting machine.After grinding, polishing and 4% nitric acid alcohol etching the polished surface,and the results using metallographic microscope to observe are shown in Fig. 8.9. 500 times 1000 times Fig.8 Simulation rolling of No.7 sample By metallographic examination, a large number of inclusions are found in 7 mm plate No.7 sample, but the inclusions distribute more uniformly,particles are not very large,generally located at the grain boundaries. banded structure appears in 4mm sheet No.6 sample,and at high magnification(1000times),the shape is slightly different with No.7(presence of the sheet-like martensite, shown in the black stripes similar to the thin line portion). 500 times 1000 times Fig.9 Simulation rolling of No.6 sample Contrasting sample 6 with 7, we can only find the thickness difference in rolling process, namely, the difference of rolling reduction
The high reduction can make the organization more compact, and easily allow higher level of aggregation of element, easily form gathering of inclusions. when there are inclusions in steel, segregation area and inclusion will be band-like distribution along the direction of deformation and form a banded structure when it is cooled.
Conclusions For HG70C, the higher the heating temperature in the hot rolling process is and the smaller the amount of reduction is,the higher the yield strength and tensile strength of the slab will be, when slab reduction quantity is large, there will be generated banded structure containing martensite, water cooling is conducive to the formation of martensite, the formation of the belt-like tissue has certain relationship with the formation of martensite[6],martensite brittleness and poor elongation characteristics will reduce the yield and tensile strength of the slab.

It is shown that the oxidation-reduction reactions and adsorption processes proceed by the exposed of PED in electroerosive reactors.
Water solutions of organic substances are involved differently in oxidation-reduction reactions and adsorption processes that are defined by the nature of MB, furatsilin and eosin.
Conversion of organic dye at different operation step Operation step of solution MB Furatsilin Eosin conversion of organic dye, [%] PED 38 44 36 Sorption on the iron loading 21 42 64 The post-effect after IED 21 18 - Total conversion 59 86 100 The data in table 2 shows that direct interaction of the PED on organic molecules does not practically depend on properties of organic molecules.
The post-effect of MB is 21% and it is defined by further oxidation-reduction reactions after finishing of PED.
According to paper [3] the coloring of MB changes as a result of oxidation-reduction reactions and it proceeds according to the scheme (1)

Instrumentation and Data Measurement This section will described the used of instruments in the experimental engine.
As seem in Figure 2, the use of CNG as a fuel under all engine speed and engine loads depicts the reduction of smoke opacity.
Conversely, the reduction of engine performance is expected causes by the lower calorific value of CNG, improper air-fuel mixture, ignition timing and combustion variations.
However, the low calorific values of the CNG fuel was expected to be the strong influence on reduction engine performances.
This reduction was expected and agreed by the previous literatures.

When varying the die semi-angle and reduction degree, drawing will result in steel rod with a hardened surface layer of the set thickness.
To study how reduction influences steel properties, a final diameter was varied from 37 mm to 35 mm.
To obtain required accuracy of data transfer between scale levels, 2 places with high concentration of elements (0.1x0.1 mm with 100 elements) were located in mesh of the rod cross section.
The difference between the test data and the simulation results was less than 6% for the surface layer and less than 2% for the central layer.
The drawing reduction degree has relatively less impact on the transformation and distribution of the martensite.

In particular, experimental data regarding load stroke curves and foam density were used to optimize the material description.
In particular, experimental data regarding load stroke curves and foam density were used to optimize the material description.
A fitting of the data concerning the specimens having a thickness equal to 15 mm was also carried out.
The exponential law resulted to fit quite well the data (R2 = 0.968).
Material flow stress data were selected from Deform database.

In order to compare output data with goal data, the mean squared error (MSE) in the feedforward backpropagation method was determined (Equation 5).
Eq. 5 where N is the number of observed data points and “ti” and “ai” denote the goal data and anticipated outputs, respectively [41, 42].
The produced nanoparticles were found to be mostly made of zero-valent iron (Fe0), according to the XRD data.
For various sets of input data, the constructed ANN may then be used to forecast heptachlor epoxide removal efficiency.
The model was significant to characterize the link between operating circumstances and removal%, according to RSM data.

EvaluationEvaluationEvaluationEvaluation ofofofof thethethethe WindWindWindWind PowerPowerPowerPower ProjectProjectProjectProject Investment Investment Investment Investment BasedBasedBasedBased onononon BPBPBPBP NeuralNeuralNeuralNeural NetworksNetworksNetworksNetworks Wei Li1, a, Shichao Li2,b, Dan Wang3,c 1College of Business Administration,North China Electric Power University,Baoding, China 2College of Business Administration,North China Electric Power University,Baoding, China 3College of Business Administration,North China Electric Power University,Baoding, China ahd11111@126.com, bwangdansky8615@tom.com, cmaggie8992@tom.com KeywordsKeywordsKeywordsKeywords::::Wind power project investment; Risk Evaluation;BP neural network AbstractAbstractAbstractAbstract.With the rapid development of the society, more and more countries have been increasingly optimistic about wind power projects because of its advantages, such as nonpolluting, renewable, energy-saving and emission reduction
However, it's non-polluting, renewable, and have made outstanding contributions to the energy saving and emission reduction.
ApplicationApplicationApplicationApplication ofofofof BPBPBPBP neuralneuralneuralneural networnetwornetwornetwor TABLE 1 EXPE RT EVALUATION FORM Take the previous nine wind power plant datas as test data for BP neural network, and the mean value from rating form as the results of expert evaluation, to analyze the data of the three wind farms, and then get a conclusion that error of network analysis agree with the results of expert evaluation, so it meet the requirements.

The calculation results showed that the band gap was reduced with the number of Si atoms in POSS increased, but the presence of hydroxyl could make a significant reduction of them.
The reduction of band gap would lower the frequency of the incident light that was absorbed by the POSS and heighten the wavelength of the absorbed incident light, which was proved by the calculation results of absorption coefficient.
Such reduction was discussed in comparison with the vinyltri- methoxysilane hybrid systems.
Characteristics such as the width of the valence band, the energy gap in insulators and the number and intensity of the main features are helpful in qualitatively interpreting experimental spectroscopic data.
The calculation results showed that the band gap was reduced with the number of Si atoms in POSS increased, but the presence of hydroxyl could make a significant reduction of them.

This paper developed feasible AID algorithms based on resampling methods to process imbalanced traffic data.
Data to be removed can be selected in a random way or, more fruitfully, according to particular criteria.
Case Study with I-880 Real Data Data description.I-880 real traffic data in California is selected in experiments to verify the performance of proposed algorithms.
Refinements in data quality promise to improve incident detection performance, therefore, develop feasible methods for data cleaning that available to AID is worthy of further research.
[11] Wilson, D., Randall and Martinez, Reduction Techniques for Instance-Based Learning Algorithms.