Search results

Manifold learning has made many successful applications in the fields of dimensionality reduction, pattern recognition, and data visualization.
In order to distinguish one particular class from others, we extract various kinds of features as much as possible from the huge amount of data, which will result in large amount of high-dimensional data sets, thus there will be the curse-of-dimensionality.
Broadly speaking, feature selection and dimensionality reduction are two feasible ways to deal with the aforementioned challenges[1].
Based on two-fold cross validation, we compare the performance of HMM with other classical dimension reduction techniques.
[5] S.Roweis and L.Saul,“Nonlinear Dimensionality Reduction by locally linear embedding,” Science, 2000, vol.290

Fig. 1(b) shows the effect of different chloride concentration on NB reduction by SM-NZVI.
Fig. 1(d) shows the effect nitrate on NB reduction with SM-NZVI.
The data showed a limited effect on NB reduction was acquired at different NO3- dosage, and the effect did not trend to be expanded significantly during the reaction.
Fig. 2(b) shows the effect of Ca2+ on NB reduction.
It means that the enhancement of NB reduction is due to Cl- but not Ca2+.

We also attribute this reduction of SF due to the surface etching prior to the drift layer growth.
Basal plane dislocation reduction is observed with a reduction in the interruption temperature (left).
SIMS data indicates etching takes place during the interruption step (Fig. 4).
Interruption at a higher temperature (Experiment-1 in Fig. 2), did not show significant reduction in BPD.
However, when the temperature was lowered further (Experiment-2, Fig.2), a significant reduction in BPD was observed.

Study on Excess Sludge Reduction in Quiescent Condition Min Xie1,2,a, Cong Huang1,b, Xiaobo Liu1,c, 1School of Hydraulic Engineering , Changsha University of Science & Technology, Changsha 410004, China; 2Research Center for Eco-Environmental Science, Chinese Academy of Sciences, Beijing 100085 ahailun301@gmail.com,b648740240@qq.com , c 46333876@qq.com Key word: excess sludge; lysis; sludge reduction; Abstract: The cell dissolved effect of acid and alkali lysis on excess sludge, as well as the influence factors including pH, temperature, time and other factors were examined in this work.
Therefore, the sludge reduction efficiency was remarkable.
The acid and alkali pretreatment on dissolving cell of excess sludge was studied, according to the release of SCOD, soluble protein, soluble total sugar and organic matter to grasp the affecting factors of excess sludge reduction of the target value.
Table 1 The characteristics of excess sludge Analytic Index pH SCOD（mg/L) SS（mg/L) VSS（mg/L) Data 6.82 79.712 8600 6450 Experimental Methods The sludge was put into a plug jar with acid and alkali solution to adjust the pH to the set value, temperature and time, and then the indicators were tested.

The ion induced-reduction eliminates the need for high temperatures typically required for the reduction processes.
The XPS spectra of the as-grown sample is characterized by a doublet with the W 4f7/2 component at a binding energy (BE) of 35.6 eV and an energy separation between the W 4f7/2 and W 4f5/2 peaks of 2.2 eV, which agrees well with the data for pure WO3 (W6+ ionic state of tungsten) from the literature [31], [36], [37].
Closed circles represent experimental data and solid lines the fitting doublets of mixed G-L functions.
These results align qualitatively with previous data obtained from WO₃ samples bombarded with Ar+ ions [36], [37].
Closed circles represent experimental data and solid lines the product of mixed G-L functions.

Review of the Application of Water Pinch Technology in Water-saving and Emission Reduction Lihong Meng1, a, Qi Qiao1 and Jingyang Liu1 1 Chinese Research Academy of Environmental Sciences, Beijing 100012, China ayuduanhongri@163.com Keywords: water pinch, water-saving, emission reduction Abstract.
And its application on the water saving emission reduction of other industries is less.
Taking a petrochemical enterprise for example, through applications of software on analysis of water data of the maximum impurity concentration of system exports, optimized system with conventional analysis and system with water networks optimal design software, we could see that use of water pinch technology optimization saved water by 48.7% than original design 14.6% than conventional optimization technology.
Zhu Shenhua[18] compared the system water data of regular analysis optimization and water pinch technology optimization with that of original design, and found that water-saving rate of routine analysis optimization and water-pinch technology optimization were 17.62% and 31.03%, respectively.
Based on experience or data analysis solution as well as the parameter mentioned above, the water network is designed which may have deviation with regeneration concentration in actual operation.

The test proved the drag reduction effect of soft dolphin skin.
Each test data is average values after test three times. 3.1 NACA4412 Airfoil The drag reduction rate, lift increase rate and lift-drag ratio change rate of airfoil are calculated by the following formulae: （1） （2） （3） Fig.8 is the relation of inflow velocity and drag for NACA4412 airfoil, and Fig.9 is the relation of inflow velocity and drag reduction rate for NACA4412 airfoil with bionic soft surface.
Turbulent drag reduction using soft surface[J].
Test Research on Drag Reduction of Soft Wall [J].
Test on Drag Reduction of Soft Surface and Research of Drag Reduction Mechanism.

Feature selection has become the focus of research areas of applications with high dimensional data.
Introduction Rapid technological developments in Computer Science have resulted in increasing quantities of data, making many of the classical data analysis tools unavailable.
NMF[3] is a good method for dimensionality reduction which has been used in many fields,such as text mining, face recognition, microarray data analysis and so on.It is always considered to be an unsupervised learning algorithms for feature extraction.This paper is to use it for feature selection not for feature extraction.
In this paper,we propose a new feature selection algorithm adopting a two-step strategy for high-dimensional data .The first step is to use improved NMF to compress the great number of data samples to several groups of vectors which can be regarded as the basis of each category in the dataset.
Since relatively few basis vectors are used to represent a lot of data vectors, good approximation can only be achieved if the basis vectors find out latent structure in the data.

Hence, the reduction ratio of gear reducer is required to be higher and higher. 3K type and 2K-2H type planetary gear trains can be designed to have high reduction ratios.
This paper proposes 2K type planetarygear train with high reduction ratio.
(a) Reduction ratio Rr=40 (b) Reduction ratio Rr=50 (c) Reduction ratio Rr=100 Fig.4. 2K type planetary gear reducers with simple planet gears The Gear Data for 2K type Planetary Simple Gear Reducer In follows, the 2K typeplanetary simple gear reducer shown in Fig. 4(a) will be the example to illustrate how to design the two ring gears with different shift coefficient to engage the same planet gear.
Table 1 show the gear data of 2K type planetary simple gear reducer shown in Fig.4(a).
Table 1, Gear data of planetary simple gear reducer shown in Fig. 4(a) Planet gear 3 Ring gear 4 Ring gear 5 Teeth No. 16 38 40 Normal Module (mm) 2.0 2.0 2.0 Shift coefficient (mm) 0.2 0.58 -0.18 Pitch Dia.

The data are collected in time-snapshots denoted by ηi= η(ti,w1)…η(ti,wn)∈Rn, i=1,…,N (1) One looks for a set of orthonormal basis vectors uj∈Rn, j=1,2,…,n , such that ηi=j=1nγjiuj , with i=1, 2,…,N , that is, η1…ηN=u1…unγ11…γ1N⋮⋱⋮γn1…γnN, U*U=In (2) where In∈Rn x n is the identity matrix, Η=η1…ηN, U=u1…un and * is the (conjugate) transpose operator.
The uj are referred to as empirical eigenfunctions or principal direction of the “cloud” of data ηi.
As a second step, a reduction of n is also investigated.
Up to now local linearizations were adopted to support the model order reduction.
Model Order Reduction: Theory, Research Aspects and Applications, Springer, 2008