Estimating the parameters in regression with uniformly distributed errors
- 1 December 1986
- journal article
- research article
- Published by Taylor & Francis in Journal of Statistical Computation and Simulation
- Vol. 26 (3-4) , 269-281
- https://doi.org/10.1080/00949658608810965
Abstract
In this paper we develop maximum likelihood estimators of the intercept and the slope in linear regression where the error term is uniformly distributed. We prove that these estimators are unbiased and we show, theoretically and via simulation, that their variances are in most practical cases smaller than the variances of the corresponding LSEKeywords
This publication has 8 references indexed in Scilit:
- An algorithm for discrete chebychev curve fitting for the simple model using a dual linear programming approachCommunications in Statistics - Simulation and Computation, 1984
- Least absolute value and chebychev estimation utilizing least squares resultsMathematical Programming, 1982
- A dual method for discrete Chebychev curve fittingMathematical Programming, 1980
- Using the least squares estimator in Chebyshev estimationCommunications in Statistics - Simulation and Computation, 1980
- On L1 and Chebyshev estimationMathematical Programming, 1973
- Two Linear Programming Algorithms for Unbiased Estimation of Linear ModelsJournal of the American Statistical Association, 1973
- A Note on the Efficiency of Least-Squares EstimatesJournal of the Royal Statistical Society Series B: Statistical Methodology, 1968
- Topics in the Investigation of Linear Relations Fitted by the Method of Least SquaresJournal of the Royal Statistical Society Series B: Statistical Methodology, 1967