Search results

So the set of data contains 150 examples with 4 dimensions and 3 classes.
Fig. 1 shows the results of data clustering based on KDA method adopting the radial basis function (RBF) kernel as Eq.4.
Comparison with KPCA KPCA, which is similar to KDA, is�Rne of the powerful techniques for extracting features from high dimensional data sets.
Data representation means that compressing data along the principal directions with little loss of structural information.
They also demonstrate the superiority of KDA for iris data classification and gear fault classification over KPCA.

Simultaneous reduction of wear and corrosion of titanium, magnesium and zirconium alloys by surface plasma electrolytic oxidation treatment Nykyforchyn H.M.a,Agarwala V.S. b , Klapkiv M.D., Posuvailo V.M. c Karpenko Physico-Mechanical Institute, 5, Naukova Str., Lviv, Ukraine a nykyfor@ipm.lviv.ua, c posuvaylo@ipm.lviv.ua U.S.
These data are the evidence of the action of corrosion factor in the course of cavitation fracture of oxide-ceramic coatings.
Some prospects for further PEO method development Summarized data of the main properties (thickness and hardness) of PEO coatings on the different alloys are presented in Table 5.
Data comparison clearly indicates the advantages of PEO treatment especially by aluminum alloys.

The same results were concluded from the molecular modelling data obtained from Chemcraft software.
The SEM-EDS result shows there is a significant reduction in the concentration of silica after modification of biochar.
The molecular composition of the BC-01 was calculated from the characterization data and a basic molecular structure of BC-01 was constructed using Avogadro software as in Figure 9a.
According to the characterization data of FeBC-01, most O-H and Si-O groups were replaced by Fe-O bonds.
The sponsors had no involvement in the methodology of research, data analysis and interpretation.

This study conducts the measurement of the outer shape of existing complex model and puts it through different steps such as data points acquisition, surface filling, and surface fitting.
Contact probe, using either continuous-contact scanning or single-point probing, acquires the height variation data of the third axis by touching the surface profile of the object.
The data file is then converted for the use of ANSYS Workbench.
Both cases show significant reduction.
In the last step, the data of the modified model is put into the rapid prototyping machine to produce a physical model which is shown in Fig. 5d.

Document and data collecting At this step, the preliminary data should be taken including, building/factory characteristic, habits and procedure of daily operation, list of energy using equipment, electricity and other energy bill, marking/nameplate from equipment, logsheet or any recording data regarding to energy consumption. 3.
Meanwhile, the data collected from the observation were analyzed and used as reference for calculation. 5.
Reporting and presentation Finally, the whole data will be written in a walkthrough energy audit report and will be presented.
Temperature, Humidity and Light intensity Data No Room Area (m2) Occupation No Room T ( C ) RH (%) Lux 1 Sampoerna corner 96 15 1 Sampoerna corner 26,7 41 226 2 R.
The data resulted from energy audit shows that heat transfer coefficient for walls is 0.65 W/m2K, and for floor is 0.5 W/m2K.

However this approach may contain high amount of uncertainty due to the arbitrariness of the data and the distributions chosen.
So most of fitted distributions match the bulk of the data at lower value, while the tail is difficult to fit which is crucial for reliable long-term design value estimation.
However, the extreme value distribution by PDEM could capture the trend in the tail of the data acceptably, especially when compared with the result of SMC shown in Fig. 3(b).
It is reasonable that the three-parameter Weibull distribution preserves the first three moments of the data.
Sutherland, Probabilistic analysis of list data for the estimation of extreme design loads for wind turbine components, 2003 ASME Wind Energy Symposium, 2003, p.0865

Data from existing literature involving acid emersion of BFRP were collected and modelled using RSM to present an overview of the degradation behavior of BFRP.
This research focuses on modelling the durability of BFRP bars after exposure to the acidic environment using statistical and data analysis as shown in Table 1.
Meanwhile, certain literature conducted testing on tensile strength but not elastic modulus, hence the missing data would be taken into consideration in RSM modelling.
Flow of study Data Collection and Curation 19 data from six previous studies are used in RSM analysis as shown in Table 1.
For the equation, the experimental and predicted tensile strength retention values are tabulated. 1 out of 19 sets of data shows large residual and unusual compared to the expected results. 94.74% of accuracy indicates that the experimental data collected from previous research are reasonable and suitable for the RSM model to produce predicted data.

The new data about the main physical effects in the treatment zone is received by optical diagnostics of the process.
It is assumed that reduction of impact of this effect shall have positive effect at microstructure of single tracks resulting in reduction of residual pore and crack formation in article and this supposes further study of possibilities of SLM method manufacturing of articles at conditions matching “the second” process window (Р=130…150 Wt, V=40…70mm/sec).
At small scanning speed (Fig.8, a) it is also observed reduction of granule ejection speed and this is associated with level of energy in laser beam.

After the paper machine drying section completed the transformation, the actual operating cases that using adjustable steam jet heat pump Further improved the machine speed, machine production increased by 15% -20%, 10% reduction in steam consumption Per ton of paper.
Data was listed in the table 1 as follows.
quantitative Paper machine speed (m/min) Percentage improvement % Before transformation After transformation 200 g/m2 480 520 1.08 230 g/m2 430 500 1.16 250 g/m2 380 460 1.21 Table .1 The data of paper machine drying section transformation Conclusion Adjustable pump for the large cardboard drying of one factory applied example shows that using adjustable steam jet heat pump to replace the unadjustable steam jet heat pump in paper machine drying could ensure heat pump system operation efficiently and steadily, it overcame the shortcomings of the unadjustable heat pump，such as: deviating from there originally working condition, decreasing the jet coefficient and power capability, limiting the paper machine’s speed to increase in the adjustment process of the load change.

The final aim of this three and half-year duration program is to provide data and models that can be used in finite elements calculations, for the design and the working simulation of industrial systems.
The associated volume reduction (about 4%) is not sufficient to explain this sudden deformation increase.
It should rapidly lead to the collection of numerous data, the compilation of which will improve the knowledge of refractories and allow numerical simulations of refractorised industrial structures.