ley de benford
Post on 01-Jan-2016
59 Views
Preview:
DESCRIPTION
TRANSCRIPT
Primer dígito significativo
299.959 0,0174
mxd
mmxx n
)(
101,10,
2,99959 ·105
[2,99959] = 21,74 ·10-2 [1,74] = 1
Las barras negras representan las frecuencias de aparición como primer dígito significativo (d = 1,2,3,...,9) en una lista de N = 201 constantes físicas.
En barras blancas aparecen las frecuencias de aparición como primer dígito de los números 1 a 9 en el tamaño en bytes de N = 1.295.777 ficheros.
Leading digit Probability
1 30.1 %
2 17.6 %
3 12.5 %
4 9.7 %
5 7.9 %
6 6.7 %
7 5.8 %
8 5.1 %
9 4.6 %
Note on the frequency of use of the different digits in natural numbers. Amer. J. Math. 4 (1881) 39-40.
Simon Newcomb (1835-1909).
d
ddP
1log)(
The law of anomalous numbers.Proc. Am. Philos. Soc. 78 (1938) 551-538.
Frank Benford
Probable Error
10114.74.95.16.48.09.412.418.530.6Average
4184.14.87.26.56.79.415.718.627.0Death Rate
9005.57.16.88.88.510.012.016.025.3 ,
3425.05.05.66.48.58.812.619.228.9Addresses
11655.44.75.27.06.68.714.117.331.0Blackbody
14583.05.64.96.47.49.812.617.632.7Am. League
7074.85.85.17.48.19.014.417.527.9X-Ray Volts
7413.15.54.75.59.810.110.118.832.4Cost Data
3084.24.95.56.57.17.512.418.533.4Reader's Digest
5605.67.37.08.48.37.514.314.826.8Design
50008.98.07.26.86.66.89.720.325.7 ,
915.54.43.34.46.64.45.518.747.2Atomic Wgt.
1591.92.55.05.08.212.613.823.927.1Drainage
18003.22.84.15.16.710.815.425.226.7Mol. Wgt.
6903.65.15.17.08.110.811.918.430.0H.P. Lost
7034.74.45.76.48.39.812.818.329.6Pressure
13894.14.83.24.110.614.616.218.424.0Specific Heat
1005.05.06.06.08.010.012.018.030.0Newspapers
10410.62.91.05.810.68.64.814.441.3Constants
32592.23.74.16.27.28.114.220.433.9Population
3355.14.25.58.67.211.310.716.431.0Rivers, Area
Sampls987654321Title
Las barras representan las frecuencias de aparición como primer dígito de los números 10 a 99 en los N = 1.295.777 ficheros medidos. La línea continua representa la ley de Benford generalizada para dos dígitos.
Invarianza de base y de escala en la densidad de probabilidadTheodore Hill
Invarianza de escala Invarianza de base
No toda lista de números que cumple la Ley de Benford proviene de una distribución invariante de escala. Pero seguro que es invariante de base.
d
dLn
d
dLn
dLndLndNN
k
k
kkd
d
k
k
1
10
)1(10
10)1(10)1(10
10
1
Para una lista de números que siga una distribución de probabilidad en forma de ley de potencias N-1, tendremos que la probabilidad del primer dígito significativo es independiente de la década y sigue la ley de Benford:
d
ddP
1log)(
Normalizando:
The demonstration of Benford’s Law (and also for the distribution of the second
digit) was done in 1996 by Professor Theodore Hill (School of Mathematics, Center for Applied Probability, Georgia Institute of Technology) in his
article: “A Statistical Derivation of the Significant‐Digit law”. Hill later showed there was a kind of central limit theorem that applied to a wide variety of distributions--that combinations of distributions tend towards the distribution predicted by Benford’s law even when the original distributions do not [Hill1996].
top related