Search results

This makes it difficult to reconstruct the data.
Additionally, LPP and OLPP are unsupervised dimensionality reduction methods which do not take the label information into account.
Defining affinity matrix is as follows (1) Where represents the neighborhood relation between data sample and .
Table 1 Recognition rates on the ORL face database (%) methods training samples/class 3 4 5 6 7 ONPDA 89.61 93.33 95.90 97.38 97.50 KONPDA 90.75 94.17 96.10 97.75 98.00 KLDA 89.27 92.67 94.70 96.00 96.83 KPCA 87.18 90.17 92.75 95.12 96.25 From Table 1, we find that KONPDA is the most efficient dimensionality reduction method, and is much more efficient than with ONPDA and kernel methods.
Conclusions In this paper, a novel dimensionality reduction approach KONPDA is developed.

In this paper, the model of the fixing plank which was shown in fig.1(a) was established in Solidworks , and the modeling process was according to the original designing data strictly.
After that, the engineering data and the boundary conditions should be set as described above.
Set the target reduction of the optimization to 20% and 38% separately, and then run the order.
Fig.4 shows the optimization result when set the target reduction to 20%.
Fig.5 shows the optimization result when set the target reduction to 38%.

In order to reduce the pollution caused by urban waste and reuse the resources, requirements of harmlessness, reduction in amount and recycle are made in the field of municipal waste disposal.
Due to the excellent reduction effect of the waste incineration technology along with its advantages of recycle and safe treatment, this kind of technology developed rapidly and took up a large proportion in the garbage disposal methods.
Therefore, the article proposes a centralized source classification approach to achieve targets of reduction, recycle and safe treatment in waste management.
Artificial neural network is inspired by biological neural network system and the application of computers to simulate its calculation function, which can effectively deal with nonlinear, incomplete data with good reliability and fault tolerance.[1] Neural network can be used to recognize the garbage in the video to achieve the automatic classification of garbage. 2 Pretreatment and characteristics of the garbage video 2.1 The read of Video Frames The method in this paper, the video camera is applied to collect the information of garbage.
When it is trained, the corresponding output weights of the data center at each hidden node will no longer change, and the neural network can enter the working state.

According to system data flow this system is constructed.
The level 0 data flow of this entire system is in Fig2.
If the data are highly correlated, there is redundant information.
Summary A text categorization system based on SVM is introduced according to system data flow.
Therefore, SVM can handle very high dimension text data.

DIT was extracted from the high frequency capacitance data by means of the Terman method [11].
The data in Fig. 4 shows DIT as a function of the energy position in the bandgap, Ec-Et.
The ConV data shown in Fig. 6 uses a constant voltage and allows projection of dielectric lifetime.
The data in Fig. 7 show the effect of a hold stress on the capacitance voltage characteristics.
The data in Fig. 8 shows Fowler Nordheim analysis of the high-field regions of J-E plots.

The calculated results show good agreement with the experimental data.
Applying the perturbation procedure similar to that in Refs. [12,13], the perturbation formulas of g factor and hyperfine structure constant for an octahedral 5d5 cluster can be expressed as: g = 2 (1 + 2kp)/3, A = P (κ/3 – 8Np /7) (1) Here kp and Np are the orbital reduction factor and the normalization factor for the π component, characteristic of covalency of the system.
From the cluster approach [14], the orbital reduction factors kp and ks can be determined by considering the ligand orbital contributions: kp = Np (1 + lpp2/2), ks = (Np Ns)1/2 [1- lpp (lps + lss w)/2] (2) Here w denotes the integral , where R is the reference impurity-ligand bond length.
Then the orbital reduction factors kπ » 0.857 and kσ » 0.461 are calculated from Eq. (2).
When the ligand orbital contributions were neglected, the theoretical results (Calc. a) are in poor agreement with the experimental data, especially the g factor and hyperfine structure constant are smaller than the observed values.

Experts pay more attention to the construction of green environmental protection of energy conservation and emissions reduction in the process of urbanization.
It had great significance for energy conservation, emissions reduction and carbon reduction [5].
Then they entitled to duty exemption or reduction for more than energy efficiency design standards or using renewable energy and green buildings.
The current domestic public construction of concrete operation data was not yet completely.

Determination of Bore Grinding Machine Parameters to Reduce Cycle Time Salakjitt Buddhachakaraa and Wipawee Tharmmaphornphilasb * Industrial Engineering Department, Faculty of Engineering, Chulalongkorn University, Bangkok, Thailand aboonsalakjitt@gmail.com, b * wipawee.t@eng.chula.ac.th Keywords: Design Of Experiment, Central Composite Design, Cycle Time Reduction Abstract.
With these settings the new cycle time is 2.76 second per piece or 7.38% reduction.
Boonsompong: Reduction of Tombstone Capacitor Problem by Six Sigma Technique: A Case Study of Printed Circuit Cable Assembly Line (2011 IEEE International Conference on Quality and Reliability, 2011)
Deeying: Determination of the Appropriate Laser Spot Welding Parameters to Reduce the Undercut Defect of the Suspension Using Split-Plot Design (The International Data Storage Technology Conference (DST-CON), 2011)

In the second part of the combustion face, reduction reactions are occurred (endothermic reaction), is called reduction zone.
, based on the obtained heat and mass balance calculations data, is carried out as to application of certain blasting interfusion types for effective underground gasification process management.
(10) where: – the average heat flow in 1 m2 of the surface rock walls channel by radiation and convection, J/m2, Sn – the surface of the coal channel, m2, τ – time for coal seam gasification to the coal channel width, s, Qi – the intensity of gasification, kg/s; Q5 – additional heat losses into the environment (unaccounted losses), made on the basis of practical data.
This greatly simplifies the data processing and allows to quickly obtain end-results with a high degree of compliance.
Definition of the Mississippian–Pennsylvanian Boundary in the Lviv–Volyn Coal Basin (Western Ukraine), Based on Palynological Data.

Such data allowed to achieve full characterization of carbide sintering.
Lack of interest in sintering single-phase carbide materials is synonymous with the lack of data on describing such process with models, which in most cases depend on measurements of sintering kinetics.
Data shown in table concern spherical grains, with no grain growth during sintering.
Measurements of sintering shrinkage under isothermal conditions in determined characteristic sintering temperatures confirmed data obtained from measurements during non-isothermal heating.
Combining the obtained data for TiC0,8 suggest, that the optimal sintering temperature is 2000°C.