Search results

(2) Data preprocessing module The module will be raw data into a neural network identification data, the processing of the raw data can be divided into three parts: the protocol analysis, packet feature selection, data processing.
According to the pre-processing module to extract the 41-dimensional feature vector for data processing, the principal component analysis (Principal Component Analysis, PCA) dimensionality reduction principal component variables 7, you can set the number of neurons in the input layer 7, the output layer neurons identified as one, can also be changed according to the actual situation of the number of neurons in the output layer
GABP network structure Simulation and Analysis (1) In this paper, experimental data from the MIT dataset KDD CUP99, these data have become in recent years, intrusion detection system evaluation standard data.
KDD CUP99 data from a random sample of about 500 sample data as a system of training and test data.
A Data Mining Framework for Building Intrusion Detection Model[C]//Proc. of 1999 IEEE Symposium on Security and Privacy.

The data are taken qualitatively for the crystal growth observation and quantitatively based on real-time temperature recording.
For each series of experiment, the measurements were repeated several times to obtain the cycling stability data of the PCM and the effect of ultrasound irradiation.
In other side, the suppression of phase separation effect in CaCl2.6H2O due to sonocrystallization treatment related to reduction of the solution’s metastable zone by the formed cavities and thus increasing the growth rate of the crystal [5].
The latter might be due to formation of cavities that lead to reduction of the solution’s metastable zone and thus increasing the growth rate of the crystal.

The surface roughness is affected by increasing of feed rate and reduction of cutting speed for 2.3 Snarled and 4.3 Snarled chip form types.
Furthermore, the analysis of cutting condition factors affecting to cutting force and surface roughness is of interest for reduction of operation time and cost to obtain proper cutting conditions.
In order to reduce the surface roughness, the reduction of feed rate is the important condition in operation.
However, reduction of both cutting force and surface roughness is also based on the obtained 4.3 Snarled chip form type obtained from the operation.
Conclusions The results obtained in this study can be drawn from evaluation of experimental data: 1) The cutting conditions that influenced on main cutting force and surface roughness in cast nylon 6 turning are different from steel turning operation. 2) Main cutting force exhibits significant increasing trend by the increasing of cutting speed for all obtained chip form types. 3) Each chip form types produced from the turning operation are based on the different initial cutting speed. 4) The surface roughness decreases by increasing of the cutting speed with production of 2.3 Snarled and 4.3 Snarled chip form types in turning operation.

A 20% addition of FA as well as a 30% addition of FA causes a reduction in fracture toughness of young concrete.
Result and discussion Based on the data recorded during the tests, graphs of rotation angle (q) vs time (t), torque (T) vs t and T vs q were plotted and values of fracture toughness were calculated.
The obtained results lead to the conclusion that 20% addition of FA as well as a 30% addition of FA causes a reduction in fracture toughness of young concrete.
For a longer curing period, 20% addition of FA caused increase of KIIIc by 4.4%, while the replacement of a part of cement with a 30% additive of FA brings about 14.1% reduction of the fracture toughness in comparison to the concrete without FA.
In the young concrete 20% addition of FA as well as a 30% addition of FA caused reduction in fracture toughness of tested concrete. 3.

In order to measure gas flows of every flow branch in the group nozzles, flow meter of type SH-1 is specifically developed, The measure system, control system, data display system, data acquisition analysis system subtly combine, the SH-1 gas flow test equipment and these measured flows data are precise, stable, good reproducibility, the errors of the measuring are less 0.5%.
The practical application proof, this valve response is keen, the stability is good, still could maintain the working pressure under the tremendous fluctuation of the designed pressure. 2.2 Measuring System The experiment table includes: plate orifice of flow, pressure sensor, differential pressure sensor, temperature sensor, humidity sensor, control valve, data acquisition clip, LED display and interface.
The equipment uses pressure-steady box to reduce the fluctuation of the gas and stabilize the results. 2.3 Data acquisition analysis system In order to digitally show the experimental results, the data acquisition card is used to input the data showed on the meters.
The basic reduction rule of the flow rate after the group nozzles weld is found out.; (4) the single-flame tube combustion tests indicated that the combustion chamber performance satisfied the designed requirements.
References [1] Zhang Baocheng, Technology of Experiment and Test of Aero-Engine, Beijing, Beijing University of Aero & Aviation Press, 2005 [2] Xu Xuefeng, Manual of Sensor and Transmitter, Beijing, Mechanical Engineering Press, 2004 [3] Xu Debiao, Data Acquisition System, Beijing, Beijing University of Aero & Aviation Press, 2000

In these applications, the needed data sets are often high dimensional and sparse, so Dimension Reduction(DR) [4][5] must be implemented before using the traditional clustering techniques.
However, DR makes the noise data and valid data closer to each other in lower dimensional space and the loss of some important correlation, which decreases the clustering quality and DUPDK effectiveness.
As for the Contextual Recommendation, because of the introduction of more context factors, the high-dimensionality and sparsity of data set turn to be more serious, the data quantity also becomes greater, so how to keep/improve the effectiveness of DUPDK based on the preservation of original correlations among input data set is a important problem.The key idea of hypergraph clustering [6] is transforming a data clustering problem into a optimization problem of hypergraph partitioning.This clustering method does not require preprocessing of DR, and uses hypergraph to keep N-ary associations among data items.Therefore it is suitable to use hypergraph partitioning for data clustering to support contextual recommendation service.
Some result reduction methods are also used to make the result more remarkable.
When the ratio of noise data is smaller, a distinct peak shows at the 35th point, while with the increase of noise data, the peak doesn’t obvious, when the ratio of noise data to whole data set reaches to 50%, the curve has not any obvious peak.

Therefore, an attempt to explain the effect of quartzite chemical composition on the electro thermal process of ferrosilicon melting was made based on a physicochemical model of a reduction process [2, 3].
The algorithm requires, as the input data, the initial composition of raw mixture, process parameters and a list of components that may appear in the equilibrium composition of each system phase.
In real conditions, the temperature Tw corresponds to the temperature of gases that are by-products in the process of silica reduction with carbon directly after leaving the furnace charge surface.
Also the theoretical yields of elements in the carbothermic reduction process were determined.

Within this context, SEA needs to keep up the pace of national strategic adjustment and takes the initiative to become an effective tool to integrate carbon reduction targets into urban decision-making process [4].
Through feasibility analysis of carbon emission reduction targets, SEA integrated with low-carbon issues aims to make low-carbon ​​recommendations in the field of urban industrial, energy, transportation, building and living consumption.
In addition, the lack of basic data is more common in China.
Therefore it is important to set scenario parameters for prediction of carbon emission reduction targets, which is a key issue to be solved for SEA integrated with low-carbon issues in the future.

The purpose of this study is to present the sound insulation performance measured in the living room and restroom spaces of real apartment buildings as basic data for further technical development to improve the insulation performance technology of inter-floor airborne sound.
Table 1 Contour of Intensity level Frequency 101㎡-1 84㎡-1 72㎡-1 Over-all 63Hz 250Hz 500Hz 1000Hz Acknowledgements This research was supported by a grant from a Strategic Research Project (A study on noise reduction solutions for adjacent households in apartments) funded by the Korea Institute of Civil Engineering and Building Technology.
Jeong, Evaluation of the Light-weight Floor Impact Sound Reduction Characteristics by Types of Resilient Material, Proc.
Kim, An Analysis on the Floor Impact Reduction of Dry Double-Floor System, Proceeding of the KSNVE. (2011)

Subsequently, 50 mmol glucose, 0.125 mol sodium hydroxide and 2 g zinc powder were added to the PVP solution under vigorous stirring to control the reduction potential and pH.
While the reduction of cupric ion was preceded, the suspension color was gradually changed from black to greenish yellow to brown to reddish brown.
However the nucleation temperature could be lowered under 100� by adding zinc powder because galvanic exchange reaction between copper and zinc was easily occurred due to their large difference of reduction potential.
X-ray diffraction pattern of prepared nanoparticles XRD pattern (Fig. 2) of the resulting product is identical to the single-phase cuprous oxide with a face-centered cubic structure and the diffraction data are in good agreement with JCPDS card of cuprous oxide (JCPDS 05-667).