Geometric Linear Regression and Geometric Relation
When a linear regression model is constructed by statistical calculation, all the data are treated without order, even if they are order data. We propose the Geometric regression and geometric relation method (GR2) to utilize the relation information inside the order of data. The GR2 transforms the order data of each variable to a curve (or geometric relation), and uses the curves to establish a geometric regression model. The prediction method using this geometric regression model is developed to give predictions. Experimental results on simulated and real datasets show that the GR2 method is effective and has lower prediction errors than traditional linear regression.
Keywordsrelations between data geometric regression geometric relation
Unable to display preview. Download preview PDF.
- 4.Yuan, Z.F., Song, S.D.: Multivariate Statistical Analysis. Science Press, Beijing (2009)Google Scholar
- 6.Chen, W.: An Introduction to Differential Manifold. High Education Press, Beijing (2001)Google Scholar
- 8.Blake, C.L., Merz, C.J.: UCI Repository of Machine Learning Databases. University of California, Irvine (1998), http://mlearn.ics.uci.edu/MLRepository.html