Approximating discrete probability distributions with. Approximating discrete probability distributions with decomposable. The representation and storing of a discrete distribution p requires then in. A method is presented to approximate optimally an n dimensional discrete probability distribution by a product of secondorder distributions, or the distribution of the firstorder tree dependence. In the domain of probabilistic neural networks the mixtures of dependence trees could help to. In probability theory and statistics chowliu tree is an efficient method for constructing a secondorder product approximation of a joint probability distribution, first. A method is presented to approximate optimally an ndimensional discrete probability distribution by a product of secondorder distributions, or the distribution of the firstorder tree dependence. The problem is to find an optimum set of n 1 first order dependence relationship among the n variables. More recently, wong and wang proposed a different product approximation. Approximating discrete probability distributions with dependence trees. Approximating discrete probability distributions with dependence trees c.
A method is presented to approximate optimally anndimensional discrete probability distribution by a product of secondorder distributions, or the distribution of the firstorder tree dependence. Quinn department of electrical and computer engineering university of illinois urbana, illinois 61801 email. Comments on approximating discrete probability distributions with dependence trees s. Abstractchow and liu considered the problem of ap proximating discrete joint distributions with dependence tree distributions where the goodness of the. Approximating discrete probability distributions with dependence trees abstract. Liu, approximating discrete probability distributions with dependence trees, ieee transactions on information theory, vol. Probability density function pdf is a statistical expression that defines a probability distribution for a continuous random variable as. For example in the figure above shows that for a pixel x k, which pixels are dependent on the others. Liuapproximating discrete probability distributions with dependence trees. Pdf approximating probability densities by mixtures of. A method is presented to approximate optimally an n dimensional discrete probability distribution by a product of secondorder distributions, or the. Poon abstractchow and liu introduced the notion of tree dependence to approximate a kth order probability distribution. Approximating discrete probability distributions with dependence. A method is presented to approximate optimally an n dimensional discrete probability distribution by a product of secondorder distributions, or the distribution.
I discrete probability distribution by a product of secondorder distributions, or the distribution of the firstorder. Approximating probability densities by mixtures of gaussian dependence trees. Dependence trees chow a dependence tree is used to apply a tree dependence to approximate probability distributions. Citeseerx document details isaac councill, lee giles, pradeep teregowda. Minimax estimation of highdimensional functionals 1.
Lectures on statistical learning theory for chowliu trees. The problem is to find an optimum set ofn 1first order dependence relationship among thenvariables. Approximating discrete probability distributions with causal dependence trees christopher j. The problem is to find an optimum set of n1 first order dependence relationship among the n variables. Coleman department of electrical and computer engineering university of illinois urbana, illinois 61801 email. A dependence tree indicates the dependence of pixels on other pixels. Optimal approximation of discrete probability distribution with kth. A firstorder dependency tree representing the product on the left.
320 1447 631 1115 824 142 748 483 110 176 330 2 472 1412 1264 608 610 81 87 1130 235 442 325 1018 205 1201 411 51 532 1196 294 539 499 214 749 671 715 534 1482 819 1088 621 329 21 635