Search results for "Probability mass function"
showing 4 items of 4 documents
Is there an absolutely continuous random variable with equal probability density and cumulative distribution functions in its support? Is it unique? …
2014
This paper inquires about the existence and uniqueness of a univariate continuous random variable for which both cumulative distribution and density functions are equal and asks about the conditions under which a possible extrapolation of the solution to the discrete case is possible. The issue is presented and solved as a problem and allows to obtain a new family of probability distributions. The different approaches followed to reach the solution could also serve to warn about some properties of density and cumulative functions that usually go unnoticed, helping to deepen the understanding of some of the weapons of the mathematical statistician’s arsenal.
Estimation of Uncertain Relations between Indeterminate Temporal Points
2000
Many database applications need to manage temporal information and sometimes to estimate relations between indeterminate temporal points. Indeterminacy means that we do not know exactly when a particular event happened. In this case, temporal points can be defined within some temporal intervals. Measurements of these intervals are not necessarily based on exactly synchronized clocks, and, therefore, possible measurement errors need to be taken into account when estimating the temporal relation between two indeterminate points. This paper presents an approach to calculate the probabilities of the basic relations (before, at the same time, and after) between any two indeterminate temporal poi…
Equivalence of the Pecka–Ponec Correlation Probability and the Statistical F Significance for MLR Models
2004
In an article of this journal Pecka and Ponec [J. Math. Chem. 27 (2000) 13] have proposed, by means of a probability calculation, a method to evaluate the statistical importance of correlations obtained from multilinear regression equations involving an arbitrary number of experimental points and parameters. Here, it is demonstrated how this probability exactly coincides with a more general concept: the confidence probability of an F distribution having the appropriate degrees of freedom.