We introduce a new method for locally adaptive histogram construction that doesn’t resort to a standard distribution and is easy to implement: the multiresolution histogram. It is based on aL 2 analysis of the mean integrated squared error with Haar wavelets and hence can be associated with a multiresolution analysis of the sample space.
Key Words and Phraseshistogram bin size selection multiresolution analysis wavelets
Unable to display preview. Download preview PDF.
- Daubechies I (1992) Ten lectures on wavelets. Society for Industrial and Applied Mathematics 61, PhiladelphiaGoogle Scholar
- Engel J (1995) Adaptive histograms based on Haar series. Unpublished manuscriptGoogle Scholar
- Gasser T, Engel J, Seifert B (1993) Nonparametric curve estimation. In: Rao (Ed.) Handbook of statistics: Computational statistics. Northholland, Amsterdam: 423–465Google Scholar
- Jawerth B, Sweldens W (1993) An overview of wavelet based multiresolution analyses. Unpublished manuscriptGoogle Scholar
- Silverman BW (1985) Density estimation for statistics and data analysis. Chapman & Hall, LondonGoogle Scholar