Downloads: 0
India | Statistics | Volume 5 Issue 7, July 2017 | Pages: 48 - 54
A New Extension of Dynamic Programming to Evaluate Linear Model
Abstract: The method Dynamic Programming (DP) can be applied on linear models which are not of full rank. The use of DP for a matrix enables to solve systems of linear equations that are unbalance and linearly dependent. This technique can be used to compute the various statistical measures such as coefficient of determination R, the t-statistic and F-statistic. These measures are very much important to check how well a model fits the data. DP technique can be used to compute the coefficients even when multicollinearity is present among the explanatory variables under the investigation.
Keywords: Dynamic Programming, Linear Models, Principle of Optimality, Least Square Problem, Non-Orthogonal.
Rating submitted successfully!
Received Comments
No approved comments available.