Search results

Because the LDA topics exists uncertainty distribution and rough set can deal with uncertain data well, so the algorithm based on rough set can improve the accuracy of topics analysis.
Rough set doesn’t need membership functions and to know much background about the data of given problem.
Data collection and LDA topic mining We collect data from www.sohu.com, www.sina.com and other websites as training and test data, and the contents cover computers, education, sports, entertainment, technology and automotive.
Since LDA topical mining is incomplete, and rough set for dealing with non-precision data has better adaptability, a topic reduction algorithm is proposed based on rough set.
Mining Incomplete Data—A Rough Set Approach[M]//Emerging Paradigms in Machine Learning.

Feature Extraction from Engine's Data Based on Rough Sets Theory WANG Feng, JIA Lixin Beijing Military Representative Bureau, Beijing, China, 100042 Email:shangmachangcun@126.com Key words: rough set; engine; feature extraction; reduction Abstract.
This paper introduced rough set theory for analyzing the reduction process data that under load in situ of different engine hours from a special vehicle, which aims to extract the characteristic parameters from the complex data that can characterize the engine technical condition.
As the data reduction of rough set method is based on the basis of discrete data tables, and the practical application of most of the data is in a continuous value.
Therefore, the data in the original decision-making system must be processed discretely.
Data in the table for decision-making can be identified using the matrix method to write programs for simple properties, and then get, for the nuclear properties, for the attribute reduction result.

Reuse SVM to signal sequences data continuation to improve EEMD.
Improved EEMD threshold noise reduction steps.
Fig.3 is the noise reduction error after four methods.
EEMD in threshold noise reduction using white noise is effectively restrain the noise, the noise reduction result is better.
Advances in Adaptive Data Analysis, Vol.1(2009),p1-41 [8] Chongfeng Cao, Shixi Yang and Jianxin Yang.

The paper fully aware of the advantages of the attribute reduction, putting forward an attribute reduction algorithm based on mutual information, by introducing the concept of information theory, and proving it’s reliability.
Flow chart The above examples can explanation the operational principle of the mutual information in attribute reduction,and it’s operation process provides reference for the attribute reduction algorithm, now the basic flow chart of the attribute reduction algorithm which based on mutual information as follows: Figure 1.
We got the potential factors by the questionnaire and expert’s advice and literature, determined the risk level based on the data of teh previous.
Then,in order to verify the efficiency of the algorithm, now take part of the data as the decision table of the information system, as shown in chart 2.
Rough Sets—Theoretical Aspects of Reasoning About Data [M].

The results of current reversal chronopotentiometry and thermodynamic data showed that both the silicon deposition and the side reaction between SiO2 and magnesium result in the loss of magnesium and low current efficiency.
It indicates that the reduction process is diffusion-controlled.
Square wave voltammogram of SiO2 reduction on a tungsten electrode.
As is shown in Fig. 5, the electrode potential slowly drifts towards more negative values, and the electrode potential suddenly jumps to more negative values corresponds to the reduction of Mg(� ) ions after a well-defined transition time τ.�� The data plotted in Fig. 6 agree with the following equation[6]: 1/2 1/2 / 4 1/2 ln RT t E E nF t τ τ = +
Chronopotentiogram of SiO2 reduction on a tungsten electrode.

Simplification of decision tables has been investigated by many authors, the current attribute reduction methods include data analysis, discernibility matrix, information entropy [4,5], etc.
References [1] Pawlak Z: Rough sets and intelligent data analysis, Information Sciences (2002), p. 147:1-12
[3] Pawlak Z: Rough sets: Theoretical Aspects of Reasoning about Data, Warsaw (1991), p. x
[4] Yuqing Peng, GuoXi Xiao, and Xin Yang: Data Structure Algorithm Animation Demo implementation of CAI software, Journal of Continue Education of Hebei University of Technology, vol. 15(Mar. 2000), p.1-4
[9] Pawlak Z: Rough sets: Theoretical Aspects of Reasoning about Data, Warsaw (1991), p. 60

At present, domestic and international usually use drag reduction rate as indicator. 1.1 Drag Reduction Rate Calculate the drag reduction rate by the friction pressure drop after DRA added to the liquid in the indoor test section of loop.
Then evaluate the effect of DRA by the drag reduction rate.
The reduction of pipe flow resistance, in fact, is the reduction of friction coefficient.
Then, according to equation (1-3), we can calculate flowrate-enhancing rate as equation (1-5) under the conditions of friction pressure drop after DRA added is and flow rate of basic data test condition is
Including a fluid pressure buffer tank, pump, valves, fixed-length test pipeline loop system． 1-Pressure level impulse 2-Return tank 3-Dilution mixing tank 4-Nitrogen tank 5-gear pump 6-bottom valve of buffer tank 7-vent pipe 8-flowmeter D,C,B-pressure transducer Figure 2-1 Drag reduction rate loop test device Experimental Results and Analysis 3.1 The drag reduction rate and flow increase rate at difference concentration of DRA Table 3-1 DRA evaluation Experimental data concentration（ppM） 10 15 20 25 30 Drag reduction rate（%） First 45.02573 50.65915 59.35945 67.77344 66.83987 Second 43.48938 51.58114 56.90592 67.77344 66.42953 Average 44.25756 51.12015 58.13268 67.77344 66.6347 flow increase rate（%） First 39.46696 48.10752 64.97512 87.6865 84.72997 Second 37.3459 49.66902 59.68489 87.6865 83.47109 Average 38.40643 48.88827 62.33 87.6865 84.10053 Figure 3-1 Drag reduction rate of domestic 3.2 The Relationship of concentration and the drag reduction rate of DRA Figure

And then it substituted equation (8) with the data at 900-1050 ºC of Fig.2., we obtained Fig.3..
Therefore carbon gasification was controlling step of whole reduction process.
Fig.5 was obtained by substitution of the data of Fig.2 at 900-1050℃.
Fig.7 was obtained by substitution of the data of Fig.2 at 900-1050 ºC.
The reduction process of carbon-bearing pellet includes two stages and the reduction rate of the first stage is faster than the second one. 3.

Definition 4: Data sheet, C is condition attributes set, a set of all the necessary attributes in C is called the core of C, written as Core(C), the core of data sheet is public part of all reduction.
The attributes reduction of continuous fault data The diesel engine vibration data about large maintenance machinery of the decision Table shown in Table 2 shown to illustrate.
Rough Sets: Theoretical aspects of reasoning about data[M].
Rough Sets and intelligent data analysis[J].
Principle and Algorithm of Data Mining[M].

Energy conservation and emission reduction have become important issues people concern.
Table1 Current situation of energy consumption for key industries in Jilin Province Year Industry 2006 2007 2008 2009 2010 Farm and sideline food processing industry 2.62 2.25 1.99 1.70 1.14 Raw chemical materials and chemicals manufa- -cturing industry 4.48 1.47 1．62 1.20 1.04 Electricity and heating power production and supply industry 32.96 28.20 29.00 18.31 16.23 Data source: calculated from Statistical Yearbook of Jilin Province from 2006-2010 and Statistical Database for China’s Economic Development Table2 Current situation of SO2 (Ton) emission for key industries in Jilin Province Industry Year 2005 2006 2007 2008 2009 Thermal power industry 171361.77 220157.8 236382.08 186382.4 192122.6 Electricity and heating power production and supply industry 188518.09 243246.78 277980.68 210295.04 217607.09 Data source: first draft of “the 12th five-year” plans for environmental protection in Jilin Province The harm of high-energy consumption
Table3 Current situation of nitric oxide (Ton) emission for key industries in Jilin Province Industry Year 2006 2007 2008 2009 Thermal power industry 208542.6 226585.3 261908.3 288958.4 Electricity and heating power production and supply industry 288937.6 263760.67 283173.06 312779.12 Data source: first draft of “the 12th five-year” plans for environmental protection in Jilin Province Table4 Current situation of COD (Ton) emission for key industries in Jilin Province Year Industry 2005 2006 2007 2008 2009 Farm and sideline food processing industry 5480 9314 10628 11617 11278 Beverage manufacturing industry 9841 12069 9206 5354 6358 Data source: first draft of “the 12th five-year” plans for environmental protection in Jilin Province Reasons of high energy consumption and pollution of enterprises are as follows: 1) Laws and regulations on energy conservation and emission reduction of enterprises are imperfect, and legislation punishment is not enough
At present, as marketization degree of enterprise energy conservation and emission reduction improves, the contradiction between energy conservation and emission reduction and profit pursuit becomes increasingly prominent, especially when energy conservation and emission reduction can not make up the cost which has been paid, the motivation of enterprises to participate in energy conservation and emission reduction will be at a discount.
Meanings, Approaches and Strategies of Energy Conservation and Emissions Reduction.