analysis of strongly multi-collinear spectroscopic data

In chemistry, multivariate calibration methods are widely applied for investigating multi-component spectroscopic data and regression with strongly correlated data with ‘‘small observations and large predictors’’ is an issue of importance in spectroscopic analysis. interpretation of the multiple regression equation depends implicitly on the assumption that the predictor variables are not strongly interrelated. this interpretation may not be valid if there are strong linear relationships among the predictor variables and the condition of severe nonorthogonality is referred to as the problem of collinear data, or multicollinearity.it is recommended that one should be very cautious in regression analysis in the presence of multicollinearity. thus, one of the challenges for analysis of spectral data is how to deal with the colinearty problem and data analysis methods should be explored for this problem and i will focus on three questions on the collinearity:1. how does multicollinearity affect statistical inference and prediction?2. how can multicollinearity be detected?3. what can be done to resolve the difficulties associated with multicollinearity?