XI.The application of least squares
- 1 January 1931
- journal article
- research article
- Published by Taylor & Francis in Journal of Computers in Education
- Vol. 11 (68) , 146-158
- https://doi.org/10.1080/14786443109461671
Abstract
The principle of least squares is applied to the general problem of finding the most probable values of observed quantities and of parameters that are related by equations of various sorts. The entire derivation is based on the statement of the principle that the weighted sum of the squares of the residuals is to be minimized with respect to the most probable values of the observed quantities and of any parameters that enter relations between these quantities. These relations will vary in form and number from one type of problem to another. The general solution takes the form of symmetrical equations, and the term “general normal equations” is suggested for them. For a specific type of problem these reduce to a form characteristic of that type.Keywords
This publication has 3 references indexed in Scilit:
- On the determination of the parameters in an empirical formulaProceedings of the Physical Society, 1930
- Method of Least Squares and Curve FittingJournal of the Optical Society of America, 1923
- XXIV. The adjustment of observationsJournal of Computers in Education, 1920