Scribd Logo
Upload

Welcome to Scribd!

  • 78 views

    Stat Mining 53

    Uploaded by

    Adonis Huayta
    This document discusses statistical concepts related to mining engineering. It states that as sample size increases, the bias of an estimator approaches zero, making it asymptotically unbiased. An estimator of variance that uses the mean of squared deviations from the sample mean is an unbiased estimator that can be applied to any sample size. Variance is the sum of within-group and between-group variances. For an estimator to be consistent, its probability of being within a set range of the true parameter value must approach 1 as sample size increases.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Stat Mining 53

    Uploaded by

    Adonis Huayta
    78 views1 page
    This document discusses statistical concepts related to mining engineering. It states that as sample size increases, the bias of an estimator approaches zero, making it asymptotically unbiased. An estimator of variance that uses the mean of squared deviations from the sample mean is an unbiased estimator that can be applied to any sample size. Variance is the sum of within-group and between-group variances. For an estimator to be consistent, its probability of being within a set range of the true parameter value must approach 1 as sample size increases.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses statistical concepts related to mining engineering. It states that as sample size increases, the bias of an estimator approaches zero, making it asymptotically unbiased. An estimator of variance that uses the mean of squared deviations from the sample mean is an unbiased estimator that can be applied to any sample size. Variance is the sum of within-group and between-group variances. For an estimator to be consistent, its probability of being within a set range of the true parameter value must approach 1 as sample size increases.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    78 views1 page

    Stat Mining 53

    Uploaded by

    Adonis Huayta
    This document discusses statistical concepts related to mining engineering. It states that as sample size increases, the bias of an estimator approaches zero, making it asymptotically unbiased. An estimator of variance that uses the mean of squared deviations from the sample mean is an unbiased estimator that can be applied to any sample size. Variance is the sum of within-group and between-group variances. For an estimator to be consistent, its probability of being within a set range of the true parameter value must approach 1 as sample size increases.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 1

    38 Statistics for mining engineering

    Notice, that when the sample size taken is large, its bias is:

    ⎛ 1 ⎞
    lim ⎜ − σ 2 ⎟ = 0
    n→∞ ⎝ n ⎠

    Therefore, this estimator is asymptotically unbiased for this reason. It can be applied when
    the sample size is large; practically n > 30.
    By analysing the bias of the investigated estimator, a conclusion can be formulated that it
    gives estimates of the variance that are too low.
    It is easy to prove that an estimator:

    1 n
    sˆ2 ( ) ∑(
    n − 1 i =1
    i )2 (1.95)

    of the unknown variance of the general population is an unbiased one. It can be applied in
    any sample size.
    If a sample of N elements is given and it consists of k groups of size ni; i = 1, 2, …, k, the
    following equation holds:

    ∑ ∑
    k 2 k
    sn
    i =1 i i
    ( xi x )2 ni
    s2
    x si 2
    + s (x ) =
    2
    i + i =1
    (1.96)
    N N
    2
    where: si —the estimate of variance within the i-th group
    x—the arithmetic mean of the whole sample
    xi —the arithmetic mean of the i-th group
    s 2 ( xi ) —the variance between groups.
    By looking more carefully at pattern (1.96), a simple conclusion can be drawn—the more
    differentiated the population, the greater the value of its variance. The relationship (1.96) is
    called the ‘variation identity’ (Sobczyk 1996 p. 46). This pattern is useful in the calculation of
    some combined machinery systems (Czaplicki 2010a p. 230) as well as in calculations con-
    nected with the homogeneity of shovel-truck systems (ibidem p. 266).
    The second important property that must be investigated before the application of a given
    estimator is its consistency.
    It is said that estimator Tn of the parameter θ is consistent if it converges in probability
    (stochastic convergence) to the true value of the parameter, i.e. the following equation holds:

    lim P{| Tn − θ | < } = 1 ε > 0 (1.97)


    n→∞

    Looking at this relationship, one may easily come to the conclusion that enlarging the sam-
    ple size leads to a situation in which the estimates that are obtained will be closer and closer
    to the real value of the unknown parameter θ.
    Suppose one has a sequence of observations {x1, x2, …} from a normal N(μ, σ) distribu-
    tion. In order to estimate the unknown expected value μ, one uses the sample mean deter-
    mined by formula (1.93). Now assume that every element of the sample is a random variable.
    If so, the estimator (1.93) becomes a random variable. Denote it by Tn. From the properties
    of the normal distribution, we know that Tn is itself normally distributed with the mean μ
    and the variance σ2/n. Equivalently, the random variable (Tn − μ)/(σ/ n ) has a standard nor-
    mal distribution. Therefore, the following relationship holds:

    P(|Tn − μ | ≥ ε ) = P[(|Tn − μ | n / σ ≥ (ε n /σ)] = 2 [1 − ΦN(ε n /σ)] → 0

    Book.indb 38 12/9/2013 12:22:48 PM

    You might also like