Search results

Thus, it is considered to take full advantage of the data information in other databases by transfer learning.
Let be the set of class data and be the number of class data.
(2) where is the covariance of class data; and are the mean vectors of class data and all data, respectively.
Yang, “A direct LDA algorithm for high-dimensional data-with application to face recognition,” Pattern Recognition, pp:2067-2070,2001
Knowledge and Data Engineering, 2010, 22(10)

The operation of tool magazine is realized by servomotor, planetary reduction gear drives chain wheel and chain to choose the tool.
So the tool changing times is 4800 in one day, equally, tool changing 10 times per minute. 14 failure data have been gotten.
According to the data in Table 2, the cumulative corrective maintenance time for each chain-type tool magazine and ATC can be calculated: T1= 7 [h], T2= 4.5 [h], T3= 6.7 [h].
It is adopted to evaluate the failure data received by experimenting on type MDH-80 chain-type tool magazine and ATC whose MTTR are calculated.
The failure data of the tool magazine without tool-changing times should be evaluated by the traditional method.

The SEM observation showed a tremendous decrease in SiC segregation and a reduction in grain size due to high energy ball milling of the precursor powders.
The XRD data were analyzed using the Williamson-Hall method [9] to estimate the crystal micro-strain and the Scherrer equation [10] to estimate average crystal grain size.
Williamson-Hall method [9] was used to analyze the XRD data to estimate the micro-strain in the particles.
The uniform distribution of SiC in ZrB2 (or HfB2) matrix and the reduction in grain size have tremendous effect on mechanical properties of the composites.

The theoretical results are found to be in general agreement with experimental data.
For MSMA, according to experimental data, we take )2 ( 1 0 −+=∆ − ξ ξ b b d eehG
Based on experimental data [2], the material parameters and threshold values of magnetic field are listed in table 1.
The theoretical results are found to be in general agreement with experimental data.
The theoretical results are found to be in general agreement with experimental data.

Table 1 Parameters of Isolation bearing Model number The first shape factor The second shape factor Effective area [S/cm2] Datum-level -pressure [N/mm2] Eigenvalue when r is 100% Vertical stiffness [KN/m] Post-yield stiffness [KN/m] Yield strength [KN] The equivalent stiffness [KN/m] Equivalent damping ratio % LRB 400 26.3 5.84 1257 10 72.4 40.1 1308 26.5 1665 Fig. 3 Bottom column arrangement plan Isolation bearings can be regarded as the composition of two nonlinear springs in the horizontal direction and one linear spring in the vertical direction.
4 6.225 8.678 7.863 7.562 7.32 10.77 Storey drift t[mm] 0 36.491 50.329 45.124 33.714 24.251 13.26 Isolation Acceleration peak[m/s2] 2.265 1.434 1.092 1.452 1.349 0.994 1.626 Storey drift [mm] 74.36 17.335 9.336 7.7 7.312 7.364 6.375 Tianjin wave Non- isolation Acceleration peak[m/s2] 4 7.43 11.57 12.092 10.025 12.642 18.87 Storey drift [mm] 0 30.938 39.595 32.788 24.065 15.486 9.094 Isolation Acceleration peak[m/s2] 2.36 2.455 2.234 2.405 2.305 2.22 2.782 Storey drift [mm] 126.772 28.138 14.133 11.356 9.914 8.214 6.775 Ninghe wave Non- isolation Acceleration peak[m/s2] 4 5.678 8.289 11.76 13.539 14.799 17.243 Storey drift [mm] 0 73.325 105.007 97.729 78.264 53.788 29.412 Isolation Acceleration peak[m/s2] 2.058 1.42 1.14 1.709 1.632 1.195 1.78 Storey drift [mm] 129.056 31.584 18.048 16.624 14.987 12.457 9.826 Through the analysis of table 2, response of acceleration and storey drift on each floor of isolation model with new-style isolation bearings have been various degrees of reduction
Non- isolation Acceleration peak[m/s2] 2.6 2.556 2.482 2.397 2.352 2.36 2.36 Storey drift [mm] 0 0.931 0.876 0.772 0.575 0.336 0.12 Isolation Acceleration peak[m/s2] 3.165 3.195 3.224 3.244 3.26 3.269 3.273 Storey drift [mm] 8.67 0.155 0.113 0.075 0.039 0.015 0.004 Tianjin wave Non- isolation Acceleration peak[m/s2] 2.6 2.588 2.576 2.563 2.554 2.554 2.542 Storey drift [mm] 0 1.836 1.655 1.3 0.87 0.44 0.138 Isolation Acceleration peak[m/s2] 1.585 1.61 1.631 1.649 1.659 1.667 1.669 Storey drift [mm] 29.448 0.526 0.314 0.205 0.126 0.049 0.009 Ninghe wave Non- isolation Acceleration peak[m/s2] 2.6 2.597 2.657 2.697 2.73 2.759 2.773 Storey drift [mm] 0 4.58 3.46 2.305 1.316 0.592 0.168 Isolation Acceleration peak[m/s2] 0.333 0.321 0.326 0.327 0.327 0.328 0.329 Storey drift [mm] 26.266 0.469 0.324 0.23 0.143 0.068 0.02 By analyzing the Table 3, response of acceleration and storey drift on each floor of isolation model with new bearing in the vertical direction have been various degrees of reduction
Conclusions By analyzing the response of isolated structure model and non-isolated structure model under the action of earthquake and get the following conclusion. 1) Compared with non-isolation model, the responses of acceleration and storey drift on each floor of isolation model with new-style bearings have been various degrees of reduction in the horizontal direction.
The example shows that new-style of isolation bearing can be isolated obviously in the horizontal direction. 2) Compared with non-isolation model, the responses of acceleration and storey drift on each floor of isolation model with new-style bearings have been various degrees of reduction in the vertical direction.

An algorithm data processing is given that obtained during by testing of standard samples in various regimes in which the quantitative composition of various materials is determined This paper describes the addition of radiation control sensors to modern spectrometer.
Reduction of the error components of the final result by analyzing the image of the interelectrode gap during the spectral analysis of metals and alloys was investigated in work [1].
This goal in the proposed device is achieved by seven sensors are connected to the spectral analysis device transmitting data to the data influencing factors processing unit.
In more detail, the paper presents the experimental data by image low-temperature plasma processing during the analyses and conclusions for improving the measurement accuracy of the quantitative composition of the identified elements.
Shakhov, Spectrometer with the data processing unit of the influencing factors, Russia Patent №134319, UPC G 01 J3/50

Based on the energy-balance fracture theory, the removal rate is calculated and it shows good agreement with experiment data.
The slurry is a kind of shear thinning fluid according to the experiment data.
According to the experiment data, the fitted value of n is 0.366, k is 4.3, the lower and upper limits of the viscosity are 0.08 Pa⋅S and 0.001Pa⋅S, respectively.
The removal rate shows good agreement with the experiment data of 24 nm/min given by Lei [2], also it is agrees with the removal rate of 40~400nm/min given by Zhao [3].
Based on the energy-balance fracture theory and the assumption that the material of asperity is hierarchically removed, the calculated removal rate shows good agreement with experiment data.

Data collection:SMART(Bruker, 2000); cell refinement: SAINT (Bruker, 2000); data reduction: SAINT; program(s) used to solve structure: SHELTXS97 (Sheldrick, 2008); program(s) used to refine structure: SHELTXL97 (Sheldrick, 2008); molecular graphics: (Farrugia, 1997) and PLATON (Spek, 2009); software used to prepare material for publication: SHELXL97 and PLATON.
Crystal and experiment data are listed in Table 1.
Crystal and experimental data Empirical formula C10 H12 N2 O3 S2 Formula weight 272.34 Temperature 293(2) K Wavelength 0.71073 Å Crystal system, space group orthorombic, Pbca Unit cell dimensions a = 8.142 Å α = 90 º.
b = 13.831 Å β = 90 º c = 21.878 Å γ = 90 º Volume 2463.7 Å3 Z, Calculated density 8, 1.468 Mg/m3 Absorption coefficient 0.430 mm-1 F(000) 1136 Theta range for data collection 1.86 to 25.48 deg.
Limiting indices -6<=h<=9, -16<=k<=15, -26<=l<=24 Reflections collected / unique 11011 / 2285 [R(int) = 0.0693] Completeness to theta = 25.48 99.9 % Refinement method Full-matrix least-squares on F2 Data / restraints / parameters 2285 / 0 / 157 Goodness-of-fit on F^2 1.064 Final R indices [I>2sigma(I)] R1 = 0.0572, wR2 = 0.1465 R indices (all data) R1 = 0.0808, wR2 = 0.1589 Largest diff. peak and hole 0.446 and -0.197 e.Å-3 Figure 2.

This idea, proposed by Friedman and Turkey in 1974, is observing data from different angles and looking for the optimum pursuit method which can reflect the data characteristic at utmost and dig data information sufficiently.
The projection pursuit method is an effective dimensionality reduction technology which is applied to the analysis and dealing with the higher dimensional, nonlinear and abnormal problems.
In essential, p-dimensional data is integrated into as the value in projected direction.
Through solving the maximum value of projection index function, we can obtain the optimal direction of projection, and fully reveal certain structure features of high dimensional data.
Table 1 shows the basic data of each stage.

Since real-time and communication amount is crucial for the wireless sensor network target tracking, the performance of target tracking in the wireless sensor network is critically depended on real-time and communication amount reduction.
But in this algorithm, large amounts of data need to be transmitted between cluster head nodes.
Furthermore, because of the particularity of the adaptive algorithm, the cluster nodes are not need to exchange large amount of data.
Distributed algorithm In order to overcome these shortcomings of the centralized tracking, the tracking strategy should use distributed dynamic clustering. 1). when the target is moving in their detection range, wake up the sensors and choose the largest as the head node; 2). all the nodes which are within 10m away from head node group cluster with the head node; 3). when the target moves out of the detection range of a node in the cluster, reselect the head node to form a new cluster; 4). measurement data and status information of the original cluster head is transferred to the new cluster head for target tracking; 5). repeating the clustering process until the target move out the wireless sensor network.
Varshney,“Target location estimation in wireless sensor networks using binary data”, In Conference on Information Sciences and Sytems, 2004