scholarly journals A Probability Density Function Generator Based on Deep Learning

Author(s):  
Chi-Hua Chen ◽  
Fangying Song ◽  
Feng-Jang Hwang ◽  
Ling Wu

To generate a probability density function (PDF) for fitting probability distributions of real data, this study proposes a deep learning method which consists of two stages: (1) a training stage for estimating the cumulative distribution function (CDF) and (2) a performing stage for predicting the corresponding PDF. The CDFs of common probability distributions can be adopted as activation functions in the hidden layers of the proposed deep learning model for learning actual cumulative probabilities, and the differential equation of trained deep learning model can be used to estimate the PDF. To evaluate the proposed method, numerical experiments with single and mixed distributions are performed. The experimental results show that the values of both CDF and PDF can be precisely estimated by the proposed method.

Author(s):  
Christophe Chesneau ◽  
Lishamol Tomy ◽  
Jiju Gillariose

AbstractThis note focuses on a new one-parameter unit probability distribution centered around the inverse cosine and power functions. A special case of this distribution has the exact inverse cosine function as a probability density function. To our knowledge, despite obvious mathematical interest, such a probability density function has never been considered in Probability and Statistics. Here, we fill this gap by pointing out the main properties of the proposed distribution, from both the theoretical and practical aspects. Specifically, we provide the analytical form expressions for its cumulative distribution function, survival function, hazard rate function, raw moments and incomplete moments. The asymptotes and shape properties of the probability density and hazard rate functions are described, as well as the skewness and kurtosis properties, revealing the flexible nature of the new distribution. In particular, it appears to be “round mesokurtic” and “left skewed”. With these features in mind, special attention is given to find empirical applications of the new distribution to real data sets. Accordingly, the proposed distribution is compared with the well-known power distribution by means of two real data sets.


Author(s):  
V. S. Huzurbazar

Let f(x, αi) be the probability density function of a distribution depending on n parameters αi(i = 1,2, …, n). Then following Jeffreys(1) we shall say that the parameters αi are orthogonal if


2021 ◽  
Author(s):  
Cemanur Aydinalp ◽  
Sulayman Joof ◽  
Mehmet Nuri Akinci ◽  
Ibrahim Akduman ◽  
Tuba Yilmaz

In the manuscript, we propose a new technique for determination of Debye parameters, representing the dielectric properties of materials, from the reflection coefficient response of open-ended coaxial probes. The method retrieves the Debye parameters using a deep learning model designed through utilization of numerically generated data. Unlike real data, using synthetically generated input and output data for training purposes provides representation of a wide variety of materials with rapid data generation. Furthermore, the proposed method provides design flexibility and can be applied to any desired probe with intended dimensions and material. Next, we experimentally verified the designed deep learning model using measured reflection coefficients when the probe was terminated with five different standard liquids, four mixtures,and a gel-like material.and compared the results with the literature. Obtained mean percent relative error was ranging from 1.21±0.06 to 10.89±0.08. Our work also presents a large-scale statistical verification of the proposed dielectric property retrieval technique.


Author(s):  
Robert J Marks II

In this Chapter, we present application of Fourier analysis to probability, random variables and stochastic processes [1089, 1097, 1387, 1329]. Arandom variable, X, is the assignment of a number to the outcome of a random experiment. We can, for example, flip a coin and assign an outcome of a heads as X = 1 and a tails X = 0. Often the number is equated to the numerical outcome of the experiment, such as the number of dots on the face of a rolled die or the measurement of a voltage in a noisy circuit. The cumulative distribution function is defined by FX(x) = Pr[X ≤ x]. (4.1) The probability density function is the derivative fX(x) = d /dxFX(x). Our treatment of random variables focuses on use of Fourier analysis. Due to this viewpoint, the development we use is unconventional and begins immediately in the next section with discussion of properties of the probability density function.


Author(s):  
Stephen Arrowsmith ◽  
Junghyun Park ◽  
Il-Young Che ◽  
Brian Stump ◽  
Gil Averbuch

Abstract Locating events with sparse observations is a challenge for which conventional seismic location techniques are not well suited. In particular, Geiger’s method and its variants do not properly capture the full uncertainty in model parameter estimates, which is characterized by the probability density function (PDF). For sparse observations, we show that this PDF can deviate significantly from the ellipsoidal form assumed in conventional methods. Furthermore, we show how combining arrival time and direction-of-arrival constraints—as can be measured by three-component polarization or array methods—can significantly improve the precision, and in some cases reduce bias, in location solutions. This article explores these issues using various types of synthetic and real data (including single-component seismic, three-component seismic, and infrasound).


1988 ◽  
Vol 31 (2) ◽  
pp. 271-283 ◽  
Author(s):  
Siegfried H. Lehnigk

We shall concern ourselves with the class of continuous, four-parameter, one-sided probability distributions which can be characterized by the probability density function (pdf) classIt depends on the four parameters: shift c ∈ R, scale b > 0, initial shape p < 1, and terminal shape β > 0. For p ≦ 0, the definition of f(x) can be completed by setting f(c) = β/bΓ(β−1)>0 if p = 0, and f(c) = 0 if p < 0. For 0 < p < 1, f(x) remains undefined at x = c; f(x)↑ + ∞ as x↓c.


Author(s):  
L A Rosa ◽  
S Nurrohmah ◽  
I Fithriani

The one parameter Lindley distribustion (theta) has been widely used in various field such as biology, technique, medical, and industries. Lindley distribution is capable for modelling data with monotone increasing hazard rate. However, in real life, there are situations where the hazard rate is not monotone. Therefore, to enhance the Lindley distribution capabilitiesfor modelling data, a modification can be used by using Alpha Power Transformed method. The result of the modification of Lindley distribution is commonly called Alpha Power Transformed Lindley distribution (APTL) distribution that has two parameters (alpha, theta). This new APTL distribution is appropriate in modelling data with decreasing or unimodal shaped of probability density function, and has hazard rates with increasing, decreasing, and upside-down bathtub shaped. The properties of the proposed distribution are discussed include probability density function, cumulative distribution function, survival function, hazard rate function, moment generating function, and rth moment. Themodel parameters are obtained using maximum likelihood method. The waiting time data is used as an illustration to describe the utility of APTL distribution.


Author(s):  
A. A. M. Mahmoud ◽  
R. M. Refaey ◽  
G. R. AL-Dayian ◽  
A. A. EL-Helbawy

In this paper, a bivariate Burr Type III distribution is constructed and some of its statistical properties such as bivariate probability density function and its marginal, joint cumulative distribution and its marginal, reliability and hazard rate functions are studied. The joint probability density function and the joint cumulative distribution are given in closed forms. The joint expectation of this distribution is proposed. The maximum likelihood estimation and prediction for a future observation are derived. Also, Bayesian estimation and prediction are considered under squared error loss function. The performance of the proposed bivariate distribution is examined using a simulation study. Finally, a data set is analyzed under the proposed distribution to illustrate its flexibility for real-life application.


FLORESTA ◽  
2003 ◽  
Vol 33 (3) ◽  
Author(s):  
Eduardo Quadros Da Silva ◽  
Sylvio Péllico Netto ◽  
Sebastião Do Amaral Machado ◽  
Carlos Roberto Sanquetta

Este trabalho tem como objetivo principal apresentar novas opções para o ajuste de distribuições de probabilidades que são utilizadas na Ciência Florestal. Alguns modelos contínuos apresentam certas distorções ao serem implementados no cálculo com dados oriundos de florestas naturais devido à grande variabilidade que se encontra nessas situações. Esse fato foi constatado principalmente quando se estudou a variável altura. Embora o modelo tenha sido construído com dados de alturas de florestas naturais, as fórmulas desenvolvidas poderão ser aplicadas para outras grandezas, principalmente se o gráfico apresentar assimetria, o que afasta a possibilidade de estudo por meio da distribuição normal. Neste estudo procurou-se mostrar que o modelo é adaptável também a dados de diâmetro e situações onde há simetria. Para a realização do trabalho inicialmente foram estudadas maneiras de modelar uma função matemática que pudesse ser transformada em função densidade de probabilidade. A função deveria assumir somente valores positivos, ser contínua e sua integral, considerando todo o intervalo real, deveria convergir para um. Foram feitas várias tentativas com funções matemáticas que, apesar de atenderem às condições de uma função densidade de probabilidade, não eram suficientemente flexíveis para se adaptar às características dos dados de uma floresta natural. Finalmente chegou-se a uma função que é definida por três sentenças, formada por um polinômio de grau n, uma curva crescente e uma curva decrescente positiva tendendo a zero com integral convergente no infinito. O polinômio explicou a maior parte dos dados e, para as classes onde este não produziu bom ajuste, foram elaboradas outras duas funções. Para os testes iniciais foram utilizados dados de alturas de Jequitibá-Rosa (Cariniana legalis), provenientes da Fazenda Reata, situada no município de Cássia, Minas Gerais. Para testar a aplicabilidade em outras situações procurou-se ajustar o modelo a dados de diâmetros e, após a aplicação do teste de Kolmogorov-Smirnov, os resultados mostraram-se satisfatórios. PROBABILITY DENSITY FUNCTION APPLICABLE TO FORESTRY Abstract The main objective of this research was to introduce new options for the fitting of probability distributions used in Forestry. Some continuous probability distributions present certain distortion when used in calculus with data from natural forests due to the high variability in such situations. This fact was especially noticed in studies of the variable height. Although the object of the study was natural forest’s height data, the developed formulas may be applied for other variables as well, especially if the resulted distribution is asymmetric which prevents the study to be made by the normal distribution curve. Before the study could be carried out, we did some work to model a mathematical function that could be changed into probability density function, that is, the functional values had to be positive, the function should be continuous, and its integral – considering the whole real interval – had to converge to one. Several attempts were made with mathematical functions that fulfilled the requirements of probability density function, but none was flexible enough to suit the data of a natural forest. Finally, a function was obtained which was defined by several sentences, formed by an n-polynomial, preferably a 5 degree increasing curve, a positive decreasing curve tending to zero, and the integral converging to the infinite. The polynomial explains most of the data; for the cases in which it fails to produce a good fitting, two other functions were created. The species used at first was Jequitibá-Rosa (Cariniana legalis), whose data came from a farm located in the municipality of Cássia, in Minas Gerais, Brazil. The total frequency observed for Jequitibá-Rosa, which was 493, was explained by the model with 492.9. The mean height achieved in the developed model 17.8 m is very close to the one that was calculated directly through observed data whose value is 18 m. An optimal adjustment was also achieved for the variance, leading to extending the research to other species and comparing the data obtained with other existing distributions.


1969 ◽  
Vol 6 (02) ◽  
pp. 442-448
Author(s):  
Lionel Weiss

Suppose Q 1 ⋆, … Q n ⋆ are independent, identically distributed random variables, each with probability density function f(x), cumulative distribution function F(x), where F(1) – F(0) = 1, f(x) is continuous in the open interval (0, 1) and continuous on the right at x = 0 and on the left at x = 1, and there exists a positive C such that f(x) &gt; C for all x in (0, l). f(0) is defined as f(0+), f(1) is defined as f(1–).


Sign in / Sign up

Export Citation Format

Share Document