CSI 5388 Functional Elements of Statistics for Machine

  • Slides: 23
Download presentation
CSI 5388: Functional Elements of Statistics for Machine Learning Part I 1

CSI 5388: Functional Elements of Statistics for Machine Learning Part I 1

Contents of the Lecture n Part I (This set of lecture notes): • Definition

Contents of the Lecture n Part I (This set of lecture notes): • Definition and Preliminaries • Hypothesis Testing: Parametric Approaches n Part II (The next set of lecture notes) • Hypothesis Testing: Non-Parametric Approaches • Power of a Test • Statistical Tests for Comparing Multiple Classifiers 2

Definitions and Preliminaries I n n n A Random Variable is a function, which

Definitions and Preliminaries I n n n A Random Variable is a function, which assigns unique numerical values to all possible outcomes of a random experiment under fixed conditions. If X takes on N values x 1, x 2, . . x. N, such that each xi є R, then, The Mean of X is The Variance is The Standard Deviation is 3

Definitions and Preliminaries II n n Sample Variance Sample Standard Deviation 4

Definitions and Preliminaries II n n Sample Variance Sample Standard Deviation 4

Hypothesis Testing n n n Generalities Sampling Distributions Procedure One- versus Two-tailed tests Parametric

Hypothesis Testing n n n Generalities Sampling Distributions Procedure One- versus Two-tailed tests Parametric approaches 5

Generalities n n Purpose: If we assume a given sampling distribution, we want to

Generalities n n Purpose: If we assume a given sampling distribution, we want to establish whether or not a sample result is representative of the sampling distribution or not. This is interesting because it helps us decide whether the results we obtained on an experiment can generalize to future data. Approaches to Hypothesis Testing: There are two different approached to hypothesis testing: Parametric and Non. Parametric approaches 6

Sampling Distributions n n Definition: The sampling distribution of a statistic (example, the mean,

Sampling Distributions n n Definition: The sampling distribution of a statistic (example, the mean, the median or any other description/summary of a data set) is the distribution of values obtained for that statistics over all possible samplings of the same size from a given population. Note: Since the populations under study are usually infinite or at least, very large, the true sampling distribution is usually unknown. Therefore, rather than finding its exact value, it will have to be estimated. Nonetheless, we can do so quite well, especially when considering the mean of the data 7

Procedure I n n Idea: If we assume a given sampling distribution, we want

Procedure I n n Idea: If we assume a given sampling distribution, we want to establish whether or not a sample result is representative of the sampling distribution or not. This is interesting because it helps us decide whether the results we obtained on an experiment can generalize to future data. Example: If a sample mean we obtain on a particular data sample is representative of the sampling distribution, then we can conclude that our data sample is representative of the whole population. If not, it means that the values in our sample are unrepresentative. (Perhaps this sample contained data that were particularly easy or particularly difficult to classify). 8

Procedure II 1. 2. 3. 4. 5. State your research hypothesis Formulate a null

Procedure II 1. 2. 3. 4. 5. State your research hypothesis Formulate a null hypothesis stating the opposite of your research hypothesis. In particular, the null hypothesis regards the relationship between the sampling statistics of the basic population and the sample result you obtained from your specific set of data. Collect your specific data and compute the statistic’s sample result on it. Calculate the probability of obtaining the sample result you obtained if the sample emanated from the data set that gave you the original sample statistic. If this probability is low, reject the null hypothesis, and state that the sample you considered does not emanate from the data set that gave you the original sample statistic. 9

One- and Two-Tailed Tests n If H 0 is expressed as an equality, then

One- and Two-Tailed Tests n If H 0 is expressed as an equality, then there are two ways to reject H 0. Either the statistic computed from your sample at hand is lower than the sampling statistics or it is higher. If you are only concerned about either lower or higher statistics, then you should perform a one-tailed test. If you are simultaneously concerned about the two ways in which H 0 can be rejected, then you should perform a two-tailed test. 10

Parametric Approaches to Hypothesis Testing n n The classical approach to hypothesis testing is

Parametric Approaches to Hypothesis Testing n n The classical approach to hypothesis testing is parametric. This means that in order to be applied, this approach makes a number of assumptions regarding the distribution of the population and the available sample. Non-parametric approaches, discussed later do not make these strong assumptions, although they do make some assumptions as well, as will be discussed there. 11

Why are Hypothesis Tests often applied to means? n n Hypothesis tests are often

Why are Hypothesis Tests often applied to means? n n Hypothesis tests are often applied to means. The reason is that unlike for other statistics, the standard deviation of the mean is known and simple to calculate. Since, without a standard deviation, hypothesis testing could not be performed (since the probability that the sample under consideration emanates from the population that is represented by the original sampling statistics is linked to this standard deviation), having access to the standard deviation is essential. 12

Why is the standard deviation of the mean easy to calculate? n Because of

Why is the standard deviation of the mean easy to calculate? n Because of the important Central Limit Theorem which states that no matter how your original population is distributed, if you use large enough samples, then the sampling distribution of the mean of these samples approaches a normal distribution. If the mean of the original population is μ and its standard deviation σ, then the mean of the sampling distribution is μ and its standard deviation σ/sqrt(N). 13

When is the sampling distribution of the mean Normal? n n The number of

When is the sampling distribution of the mean Normal? n n The number of samples necessary for the sampling distribution of the mean to approach normal depends on the distribution of the parent population. If the parent population is normal, then the sampling distribution of the mean is also normal. If the parent population is not normal, but symmetrical and uni-modal, then the sampling distribution of the mean will be normal, even for small sample sizes. If the population is very skewed, then, sample sizes of at least 30 will be required for the sampling distribution of the mean to be normal. 14

How are hypothesis tests set up? t-tests n n Hypothesis Tests are used to

How are hypothesis tests set up? t-tests n n Hypothesis Tests are used to find out whether a sample mean comes from a sampling distribution with a specified mean. We will consider: • One-sample t-tests n n μ, σ known μ, σ unknown • Two-sample t-tests n n Two-matched samples Two-independent samples 15

One-sample t-test σ known n n If σ is known, we can use the

One-sample t-test σ known n n If σ is known, we can use the central limit theorem to obtain the sampling distribution of this population’s mean (mean is μ and standard deviation is σ/sqrt(N)). Let X be the mean of our data sample, we compute z = (X – μ)/(σ/sqrt(N)) (1) We find the probability that z is as large as the value obtained from the z-table and then output this probability if we are solely interested in a one-tailed test and double it before outputting it if we are interested in a two-tailed test. If this output probability is smaller than. 05, we would reject H 0 at the. 05 level of significance. Otherwise, we would state that we have no evidence to conclude that H 0 does not hold. 16

What is the meanings and purpose of z? n n n Normal distributions can

What is the meanings and purpose of z? n n n Normal distributions can all be easily mapped into a single one, using a specific transformation. This means that, in our hypothesis tests, we can use the same information about the sampling distribution over and over (if we assume that our population is normally distributed), no matter what the mean and variance of our actual population are. Any observation can be changed into a standard score, z, with respect to mean=0 and standard deviation =1, as follows: Z = (X-mean)/sd 17

One-sample t-test σ unknown n In most situations, σ, the variance of the population

One-sample t-test σ unknown n In most situations, σ, the variance of the population is unknown. In this case, we replace σ by s, the sample standard deviation, in equation (1) yielding t = (X – μ)/(s/sqrt(N)) (2) Because s is likely to under-estimate σ, and, thus, return a t-value larger than z would have been had σ been known, it is inappropriate to use the distribution of z to accept or reject the null hypothesis. Instead, we use the Student’s t distribution, which corrects for this problem and compares t to the ttable with degree of freedom N-1. We then proceed as we did for z on the slide about σ known, above. 18

What is the meanings and purpose of t? n n n t follows the

What is the meanings and purpose of t? n n n t follows the same principle as z except for the fact that t should be used when the standard deviation is unknown. t, however, represents a family of curves rather than a single curve. The shape of the t distribution changes from sample size to sample size. As the sample size grows larger and larger, t looks more and more like a normal distribution 19

Assumption of the t-test with σ unknown n Please, note that one assumption is

Assumption of the t-test with σ unknown n Please, note that one assumption is made in the use of the t-test. That is that we assume that the sample was drawn from a normally distributed population. This is required because the derivation of t by Student was based on the assumption that the mean and variance of the population were independent, an assumption that is true in the case of a normal distribution. In practice, however, the assumption about the distribution from which the sample was drawn can be lifted whenever the sample size is sufficiently large to produce a normal sampling distribution of the mean. In general, n= 25 or 30 (number of cases in a sample) is sufficiently large. Often, it 20 can be smaller than that.

Two-sample t-tests matched samples n n Given two matched population, we want to test

Two-sample t-tests matched samples n n Given two matched population, we want to test whether the difference in means between these two populations are significant or not. We do so by looking at the difference in means, D, and variance, SD, between these two populations and comparing it to the mean of 0. We can then apply the t-test as we did above, in the case where σ was unknown. This time, we have t = (D – 0)/ (SD/sqrt(n)) (3) We use the t-table as before with a n-1 degree of freedom, and the same assumptions about the normality of the distribution. 21

Two-sample t-tests independent samples n n This time, we are interested in comparing two

Two-sample t-tests independent samples n n This time, we are interested in comparing two populations with different means and variance. The two populations are completely independent. We can, again apply the t-test, with the same conditions applying, using the formula: t= (X 1 –X 2)/ sqrt((s 12/n 1) + (s 22/n 2)) 22

Confidence Intervals n n Sample means represent point estimates of the mean parameter. Here,

Confidence Intervals n n Sample means represent point estimates of the mean parameter. Here, we are interested in interval estimates, which tell us how large or small the true value of μ could be without causing us to reject H 0, given that we ran a t-test on the mean of our sample. To calculate these intervals, we simply take the equations presented on the previous slides and express them in terms of μ, and as a function of t. We then replace t for the two-tailed value we are interested in in the t-table. This value can be positive or negative, meaning that we will obtain two values for μ: μupper and μlower. This gives us the limits of the confidence interval. The confidence interval means that μ has a certain probability (attached to the value of t chosen) to belong to this interval. The greater the size of the interval, the greater the probability that μ is included. Conversely, the smaller that interval, the smaller the probability that it is included. 23