Linear perturbation models for product quality,linearized around the nominal trajectories, are identified from process operational data using principal component regression (PCR) and partial least squares (PLS) regression.

This paper makes research on various methods of real-time flood forecasting at the Sujiahekou dam site in the Binlang River basin,including recursive least squares method,orthogonal projection method,forgetting factor method,covariance-adjusting forgetting factor method and principal component regression method.

Aiming at the online estimate of consistency of Br ion that is in oxidizing reactor in PTA unit,four soft sensing models are built based on BPNN,RBFNN, PCR and PLS.

Fault diagnosis for excavator′s hydraulic system based on principal component regression model;

Multiple regression of the related factors on the attack rate of influenza outbreak in Hangzhou city;

A method of optimizing a candidate zone to be fractured with multiple regression and neural network is put forward in the paper because there are some problems in Wuliyasi Sag, such as strong reservoir heterogeneity, poor interlayer shielding, big difference between fractured wells and it is hard to select good candidate wells.

Then we put forward the computation formula for base distortion and energy consumption of multiple regression with fortified intensity, base thickness and liquid level.

For the problem that the dependent variable has many independent variables and their sampling periods are also different, a predicting method is proposed by using synthetically the data analysis methods of support vector regression (SVR), multivariate regression and principal component analysis, etc.

The inherent laws behind the numerical data were discovered after the data were processed by the multivariate regression model and computer simulation.

The multivariate regression has two defects.