Search results

The example of value reduction based on the improved algorithm illustrated that it is an effective value reduction algorithm and an important supplement of basic value reduction algorithm.
Introduction The task of knowledge discovery in database is to find the internal links among large amount of data and finally access to interesting decision-making rules.
[2] Chan P and Stolfo S: On the Accuracy of Meta-learning for Scalable Data Mining, Journal of Intelligent Systems, vol. 8, 1997 pp.5-28
[3] Anannd S: EDM: A General Framework for Data Mining Based on Evidence Theory, Data & Knowledge Engineering, vol. 18, 1996, pp.189-223
[4] Chen M.: Data Mining: An Overview from a Database Perspective, IEEE Transactions on Knowledge and Data Engineering, vol. 8, 1996, pp.866-883

Make Huangbai criterion and sample data as weibull distribution to calculate similar.
Information reduction method description Information reduction method is based on rough set[3, 4].
Rough set theory can be regarded as a new mathematical tool for imperfect data analysis.
Rough set based data analysis starts from a data table called a decision table, columns of which are labeled by attributes, rows---by objects of interest and entries of the table are attribute values.
After calculated, the fuzzy centre data of recommended samples are(0.0694±0.0731, 0.1158±0.3044, 0.0131±0.0144); the fuzzy centre data of no-recommended samples are(0.1684±0.0983, 0.0318±0.0177, 0.0077±0.0083).

In practical engineering, equipment condition data are usually produced incrementally, so how to minimize repeated reduction on the new data set has an important significance to improve the efficiency of the diagnosis algorithm.
The resolution matrix for data expansion.
By calculating the matrix and doing reduction, we can get the reduction result after data expansion: Table 1 : Reduction results after data expansion Reduction results after data expansion Data type Changes of U1 and U2 Changes of M1 and M2 Reduction results 1) R(U)={C4,C7,C10} 2) R(U)={C3,C8,C9} 3) R(U)={C4,C7,C10} 4) R(U)={C3,C7,C10} In the examples after attributes added, here, we mainly extracted the following 4 frequency domain characteristics, which are expressed as P1, P2, P3 and P4 [7]: ,,, (4) In the formulas , is spectrum , K represent for the number of spectral lines , and K=1，2，…K .
What’s more, with the reduction results and bearing data, we can get the bearing running states and can better monitor the bearing workings.
Dang, Attribute reduction for dynamic data sets, Applied Soft Computing. 13 (2013) 676–689

In this work, the RSM, simulated in FORTRAN, is validated with hot rolling experimental data for higher reductions.
The modified model was simulated and validated with hot rolling experimental data for hot rolling schedules at low strain rates and low reductions (<10%).
The modified model was then simulated in FORTRAN and validated with hot rolling experimental data for different hot rolling schedules at high reductions (up to 22.7%).
The required input data were rolling speed, furnace temperature, initial and final height of the specimen, and specimen width.
Results and Discussion The output of the FORTRAN codes shows temperature data for AISI316 specimen with different geometrical forms.

Considering seismic data are always nonlinear and non-stationary, many people applied EEMD to noise reduction.
(2) Make EEMD to seismic data
(a) Original noise-free synthetic seismic data, (b) original noisy synthetic seismic data, (c) noise reduction by wavelet method, and (d) noise reduction by our method To further test the performance, our method is applied to a real post-stack seismic data as shown in Fig.2 (a), consisting of 100 traces with a sample interval of 2ms.
Local singular value decomposition for signal enhancement of seismic data.
Advances in Adaptive Data Analysis, 1(1), pp. 1-41, 2009.

According to the results of simulation, it shows good agreement with experimental data.
From the data in the figure, it can be seen in different operating points the error between simulation result and experimental data is very small, generally within 5%.
Comparison of simulation results with experimental data.
Comparison of simulation results with experimental data.
All these simulation results compared with experimental data shows good agreement.

Thus, it is more effective to obtain a more realistic data mined from the vast amounts of geological measure data than other methods.
The prediction model established by it shown in Figure 1, is formed by the four sub-modules: data preprocessing, attribute reduction and rule generation and prediction.
Through data cleaning to smooth the geological drilling noisy data samples, and fill the vacancies value of the data sheet, remove isolated points [8].
It can be solved by smoothing methods or replaced noise data by the average data.
In this paper, using the probability distribution law to discrete a continuous multiple condition attribute, according to data distribution between the small segment to determine a threshold value ,and it is more than the discrete segment data value, then merge it with "near" range to, if more than the threshold will be divided into a class. 4 Attribute Reductions 4.1 Distinguish Matrices Attribute reduction directly related to the quality and efficiency of data mining is the core of the algorithm.

Basic Attribute Reduction Algorithm Based on Discriminability Matrix Algorithm Description The general steps of attribute reduction based on discriminability matrix are: firstly obtain the core of attribute reduction set with discriminability matrix and then to calculate attribute reduction set with reduction algorithm.
Improved Attribute Reduction Algorithm The minimum reduction is same as others reduction of an information system, which are both NP complex problems.
References [1] Chan P and Stolfo S. : On the Accuracy of Meta-learning for Scalable Data Mining, Journal of Intelligent Systems, vol. 8, 1997 pp.5-28
[2] Anannd S. : EDM: A General Framework for Data Mining Based on Evidence Theory, Data & Knowledge Engineering, vol. 18, 1996, pp.189-223
[5] Chen M: Data Mining: An Overview from a Database Perspective, IEEE Transactions on Knowledge and Data Engineering, vol. 8, 1996, pp.866-883.

In the course of this ongoing collaboration, the organization and dissemination of information and data play a key role.
Therefore, it is necessary to introduce data filtering function during the transmission of information flow between the various departments.
It has three main functions: (1) collect related data and information of disaster from the public, media and specific experts; (2) organize useful data; (3) send the necessary final filtered metadata, data and information to its subordinate.
In all stages of the disaster management process, the applied information and communication technologies includes the internet, information security, computer databases, integrated data management, intelligent data management, audit tools, multi-agent systems, multi-language support, data filtering, mobile communication technology, wired communication technologies, and wireless communication technologies.
Research on sharing of international disaster data and information, Disaster, vol.3, pp.109-113, 2008

Based on the statistical data of Changsha-Zhuzhou-Xiangtan (CZT) urban agglomeration from the year 2005 to 2010, the environmental learning curves of sulfur dioxide emission per 10 thousand Yuan (RMB) production value and per capita gross domestic product (GDP) were established, and sulfur dioxide emission reduction potential of these cities was analyzed.
Data Source.
All data used in this paper came from Statistical Yearbook of Hunan Province, Changsha Statistical Yearbook, Zhuzhou Statistical Yearbook, Xiangtan Statistical Yearbook and the web site of Hunan Provincial Statistics Bureau.
From the data, it also can be seen that the decrease of sulfur dioxide emission per 10 thousand Yuan production value in Changsha city is the largest, and then in Zhuzhou city.
Sulfur dioxide emission reduction potential.