Principal Component Regression

In this paper, we derive a principal component regression (PCR) method for estimating the optical flow between frames of video sequences according to a pel-recursive manner. This is an easy alternative to dealing with mixtures of motion vectors due to the lack of too much prior information on their statistics (although they are supposed to be normal). The 2D motion vector estimation takes into consideration local image properties. The main advantage of the developed procedure is that no knowledge of the noise distribution is necessary. Preliminary experiments indicate that this approach provides robust estimates of the optical flow. KEY WORDS Motion estimation, principal component regression, and surveillance. 1.

The objective of this study is to compare the effects of climate change on crop yields across different regions. A Principal Component Regression (PCR) model is developed to estimate the historical relationships between weather and crop yields for corn, soybeans, cotton, and peanuts for several northern and southern U.S. states. Climate change projection data from three climate models are applied

Utilization of genetic resources in developing sustainable solutions to basic crop constraints have been suggested from time to time, but these genetic resources could not be exploited fully due to their inherent problems of large size and lack of sufficient evaluation and classification (Dahberg, 1995). Germplasm maintenance, evaluation, and characterization for economically important traits are prerequisite for genetic improvement program of any crop. This principal components approach could resolve several phenotypic ...

Generally, non-linear predictive models should be superior to linear predictive models. The objective of this study is to compare the performance of soluble solid content (SSC) prediction via Artificial Neural Network with Principal Components (PCs-ANN) and Principal Component Regression (PCR) in Visible and Shortwave Near Infrared (VIS-SWNIR) (400 - 1000 nm) spectrum. The spectra of 116 Fuji Apple samples were separated into calibration set of 84 apple samples and testing set of 32 apple samples randomly. Firstly, multiplicative scattering correction (MSC) was used to pre-process the spectra. Secondly, Principal Component Regression (PCR) was used to obtain the optimal number of principal components (PCs). Thirdly, the optimal PCs were used as the inputs of both multiple linear regression (MLR) and Artificial Neural Network (ANN) models. The results from this study showed that the predictive performance was improved significantly when PCs-ANN with two neurons was used compared to the PCR.

The Principal Component Analysis onto References Subspaces is a multivariate method to analyze two sets of quantitative variables when between the two sets exists a directional relationship. When the explicative variables are affected by multicollinearity this technique is not recommended. In literature exist many methods to resolve this problem (Ridge Regression, Principal Component Regression, Partial Least Square, Latent Root Regression), this work shows an alternative method based on simple linear regression.