EDGE ESTIMATION IN POPULATION OF PLANER GRAPHS USING SAMPLING
Keywords:
Graph, Planer Graph, Edge, Vertices(nodes), Simple Random Sampling without replacement (SRSWOR), Class, Estimator, Bias, Mean Squared Error (MSE), Optimum Choice, Confidence intervals.Abstract
Consider a population which contains graphical relationship between two variables. There are two graphs of vertices and edges, each edge contains a length value and linked with two vertices (nodes). Mean length of all edges is unknown which is a problem to explore. This paper takes into account two planer graphs in particular, one of them is under main interest and other is an auxiliary graph. A sample of some nodes is drawn by simple random sampling (SRSWOR) along with a laid down node-sampling procedure and a class of estimators is proposed to estimate the mean length of an edge of planer graph using the structure of other planer graph as an auxiliary source of information. Optimal properties of estimators are derived and results are numerically supported with the calculation of length estimates and confidence intervals.
Downloads
References
Aleksandrov, L., Djidjev, H., Guo, H.and Maheshwari, A. (2007). Regular Papers:
Partitioning planar graphs with costs and weights, Journal of Experimental
Algorithmic (JEA), 11, p. 102-113.
Cochran, W.G. (2005). Sampling Techniques, John Wiley and Sons, Fifth Edition,
New York.