1 Summarizing Performance Data Confidence Intervals Important Easy to Difficult Warning: some mathematical content.

Slides:



Advertisements
Similar presentations
Review bootstrap and permutation
Advertisements

Computational Statistics. Basic ideas  Predict values that are hard to measure irl, by using co-variables (other properties from the same measurement.
October 1999 Statistical Methods for Computer Science Marie desJardins CMSC 601 April 9, 2012 Material adapted.
CHAPTER 21 Inferential Statistical Analysis. Understanding probability The idea of probability is central to inferential statistics. It means the chance.
Sampling: Final and Initial Sample Size Determination
Statistics: Purpose, Approach, Method. The Basic Approach The basic principle behind the use of statistical tests of significance can be stated as: Compare.
Sampling Distributions (§ )
4.3 Confidence Intervals -Using our CLM assumptions, we can construct CONFIDENCE INTERVALS or CONFIDENCE INTERVAL ESTIMATES of the form: -Given a significance.
1 Summarizing Performance Data Confidence Intervals Important Easy to Difficult Warning: some mathematical content.
Model Fitting Jean-Yves Le Boudec 0. Contents 1 Virus Infection Data We would like to capture the growth of infected hosts (explanatory model) An exponential.
Regression Analysis. Unscheduled Maintenance Issue: l 36 flight squadrons l Each experiences unscheduled maintenance actions (UMAs) l UMAs costs $1000.
Statistics: Data Analysis and Presentation Fr Clinic II.
Comparing Systems Using Sample Data
The Simple Regression Model
Descriptive statistics Experiment  Data  Sample Statistics Experiment  Data  Sample Statistics Sample mean Sample mean Sample variance Sample variance.
Topic 2: Statistical Concepts and Market Returns
1 Pertemuan 06 Sebaran Normal dan Sampling Matakuliah: >K0614/ >FISIKA Tahun: >2006.
2008 Chingchun 1 Bootstrap Chingchun Huang ( 黃敬群 ) Vision Lab, NCTU.
Chapter 7 Estimation: Single Population
Tests Jean-Yves Le Boudec. Contents 1.The Neyman Pearson framework 2.Likelihood Ratio Tests 3.ANOVA 4.Asymptotic Results 5.Other Tests 1.
Statistical Methods in Computer Science Hypothesis Testing I: Treatment experiment designs Ido Dagan.
BCOR 1020 Business Statistics
Review for Exam 2 Some important themes from Chapters 6-9 Chap. 6. Significance Tests Chap. 7: Comparing Two Groups Chap. 8: Contingency Tables (Categorical.
Sample Size Determination Ziad Taib March 7, 2014.
Standard error of estimate & Confidence interval.
Bootstrapping applied to t-tests
1 Summarizing Performance Data Confidence Intervals Important Easy to Difficult Warning: some mathematical content.
Review of Statistical Inference Prepared by Vera Tabakova, East Carolina University ECON 4550 Econometrics Memorial University of Newfoundland.
© 2005 The McGraw-Hill Companies, Inc., All Rights Reserved. Chapter 12 Describing Data.
CPE 619 Simple Linear Regression Models Aleksandar Milenković The LaCASA Laboratory Electrical and Computer Engineering Department The University of Alabama.
Simple Linear Regression Models
The paired sample experiment The paired t test. Frequently one is interested in comparing the effects of two treatments (drugs, etc…) on a response variable.
Go to Index Analysis of Means Farrokh Alemi, Ph.D. Kashif Haqqi M.D.
Chapter 4 – Modeling Basic Operations and Inputs  Structural modeling: what we’ve done so far ◦ Logical aspects – entities, resources, paths, etc. 
Dan Piett STAT West Virginia University
1 CSI5388: Functional Elements of Statistics for Machine Learning Part I.
Statistics & Biology Shelly’s Super Happy Fun Times February 7, 2012 Will Herrick.
Topic 5 Statistical inference: point and interval estimate
© 2003 Prentice-Hall, Inc.Chap 6-1 Business Statistics: A First Course (3 rd Edition) Chapter 6 Sampling Distributions and Confidence Interval Estimation.
Population All members of a set which have a given characteristic. Population Data Data associated with a certain population. Population Parameter A measure.
Topics: Statistics & Experimental Design The Human Visual System Color Science Light Sources: Radiometry/Photometry Geometric Optics Tone-transfer Function.
Review of Chapters 1- 5 We review some important themes from the first 5 chapters 1.Introduction Statistics- Set of methods for collecting/analyzing data.
Continuous Probability Distributions Continuous random variable –Values from interval of numbers –Absence of gaps Continuous probability distribution –Distribution.
University of Ottawa - Bio 4118 – Applied Biostatistics © Antoine Morin and Scott Findlay 08/10/ :23 PM 1 Some basic statistical concepts, statistics.
1 Statistical Distribution Fitting Dr. Jason Merrick.
Chapter 7: Sample Variability Empirical Distribution of Sample Means.
Academic Research Academic Research Dr Kishor Bhanushali M
Statistics : Statistical Inference Krishna.V.Palem Kenneth and Audrey Kennedy Professor of Computing Department of Computer Science, Rice University 1.
Confidence Interval Estimation For statistical inference in decision making:
CPE 619 Comparing Systems Using Sample Data Aleksandar Milenković The LaCASA Laboratory Electrical and Computer Engineering Department The University of.
Statistics What is statistics? Where are statistics used?
Summarizing Risk Analysis Results To quantify the risk of an output variable, 3 properties must be estimated: A measure of central tendency (e.g. µ ) A.
© Copyright McGraw-Hill 2004
Confidence Interval Estimation For statistical inference in decision making: Chapter 9.
Statistical Inference Statistical inference is concerned with the use of sample data to make inferences about unknown population parameters. For example,
SAMPLING DISTRIBUTION OF MEANS & PROPORTIONS. SAMPLING AND SAMPLING VARIATION Sample Knowledge of students No. of red blood cells in a person Length of.
Chapter 13 Understanding research results: statistical inference.
Chapter 6: Descriptive Statistics. Learning Objectives Describe statistical measures used in descriptive statistics Compute measures of central tendency.
Review Statistical inference and test of significance.
Confidence Intervals. Point Estimate u A specific numerical value estimate of a parameter. u The best point estimate for the population mean is the sample.
Virtual University of Pakistan
Outline Sampling Measurement Descriptive Statistics:
Estimating standard error using bootstrap
Confidence Intervals Cont.
STAT 312 Chapter 7 - Statistical Intervals Based on a Single Sample
Sampling Distributions and Estimation
Statistical Methods For Engineers
Econ 3790: Business and Economics Statistics
Sampling Distributions (§ )
Introductory Statistics
Presentation transcript:

1 Summarizing Performance Data Confidence Intervals Important Easy to Difficult Warning: some mathematical content

Contents 1.Summarized data 2.Confidence Intervals 3.Independence Assumption 4.Prediction Intervals 5.Which Summarization to Use ? 2

3 1 Summarizing Performance Data How do you quantify: Central value Dispersion (Variability) old new

4 Histogram is one answer

5 ECDF allow easy comparison old new

6 Summarized Measures Median, Quantiles Median Quartiles P-quantiles Mean and standard deviation Mean Standard deviation What is the interpretation of standard deviation ? A: if data is normally distributed, with 95% probability, a new data sample lies in the interval

7 Coefficient of Variation Scale free Summarizes variability Second order For a data set with n samples Exponential distribution: CoV =1

Lorenz Curve Gap Alternative to CoV For a data set with n samples Scale free, index of unfairness 8

Jain’s Fairness Index Quantifies fairness of x; Ranges from 1: all x i equal 1/n: maximum unfairness Fairness and variability are two sides of the same coin 9

Lorenz Curve Old code, new code: is JFI larger ? Gap ? Gini’s index is also used; Def: 2 x area between diagonal and Lorenz curve More or less equivalent to Lorenz curve gap 10 Lorenz Curve gap Perfect equality (fairness)

11

Which Summarization Should One Use ? There are (too) many synthetic indices to choose from Traditional measures in engineering are standard deviation, mean and CoV In Computer Science, JFI and mean JFI is equivalent to CoV In economy, gap and Gini’s index (a variant of Lorenz curve gap) Statisticians like medians and quantiles (robust to statistical assumptions) We will come back to the issue after discussing confidence intervals 12

13 2. Confidence Interval Do not confuse with prediction interval Quantifies uncertainty about an estimation

14 mean and standard deviation quantiles

15 Confidence Intervals for Mean of Difference Mean reduction = 0 is outside the confidence intervals for mean and for median Confidence interval for median

16 Computing Confidence Intervals This is simple if we can assume that the data comes from an iid model Independent Identically Distributed

17 CI for median Is the simplest of all Robust: always true provided iid assumption holds

18

19 Confidence Interval for Median n = 31 n = 32

20 CI for mean and Standard Deviation This is another method, most commonly used method… But requires some assumptions to hold, may be misleading if they do not hold There is no exact theorem as for median and quantiles, but there are asymptotic results and a heuristic.

21 CI for mean, asymptotic case If central limit theorem holds (in practice: n is large and distribution is not “wild”)

22 Example n =100 ; 95% confidence level CI for mean: m ± 1.96 s / √n amplitude of CI decreases in compare to prediction interval

23 Normal Case Assume data comes from an iid + normal distribution Useful for very small data samples (n <30)

24 Example n =100 ; 95% confidence level CI for mean: CI for standard deviation: same as before except s instead of   for all n instead of 1.98 for n=100 In practice both (normal case and large n asymptotic) are the same if n > 30 But large n asymptotic does not require normal assumption

25 Tables in [Weber-Tables]

26 Standard Deviation: n or n-1 ?

27 Bootstrap Percentile Method A heuristic that is robust (requires only iid assumption) But be careful with heavy tail, see next but tends to underestimate CI Simple to implement with a computer Idea: use the empirical distribution in place of the theoretical (unknown) distribution For example, with confidence level = 95%: the data set is S= Do r=1 to r=999 (replay experiment) Draw n bootstrap replicates with replacement from S Compute sample mean T r Bootstrap percentile estimate is (T (25), T (975) )

28 Example: Compiler Options Does data look normal ? No Methods and give same result (n >30) Method (Bootstrap) gives same result => Asymptotic assumption valid

Confidence Interval for Fairness Index Use bootstrap if data is iid 29

30

Question Q1.1 If data is guaranteed to be come from an iid normal distribution, can you think of a better way to compute a confidence interval for Jain’s fairness index ? 31

Confidence Interval for Success Probability Problem statement: want to estimate proba of failure; observe n outcomes; no failure; confidence interval ? This a confidence interval for the mean ? (explain why) The general theory does not give good results when mean is very small 32

33

34

35

36 Take Home Message Confidence interval for median (or other quantiles) is easy to get from the Binomial distribution Requires iid No other assumption Confidence interval for the mean Requires iid And Either if data sample is normal and n is small Or data sample is not wild and n is large enough The boostrap is more robust and more general but is more than a simple formula to apply To we need to verify the assumptions

3. The Independence Assumption 37 Confidence Intervals require that we can assume that the data comes from an iid model Independent Identically Distributed How do I know if this is true ? Controlled experiments: draw factors randomly with replacement Simulation: independent replications (with random seeds) Else: we do not know – in some cases we will have methods for time series

38 What does independence mean ?

Example Pretend data is iid: CI for mean is [69; 69.8] Is this biased ? 39 dataACF

What happens if data is not iid ? If data is positively correlated Neighbouring values look similar Frequent in measurements CI is underestimated: there is less information in the data than one thinks 40

41 4. Prediction Interval CI for mean or median summarize Central value + uncertainty about it Prediction interval summarizes variability of data

42 Prediction Interval based on Order Statistic Assume data comes from an iid model Simplest and most robust result (not well known, though):

43 Prediction Interval for small n For n=39, [x min, x max ] is a prediction interval at level 95% For n <39 there is no prediction interval at level 95% with this method But there is one at level 90% for n > 18 For n = 10 we have a prediction interval [x min, x max ] at level 81%

Prediction Interval for Small n and Normal Data Sample 44

45 Re-Scaling Many results are simple if the data is normal, or close to it (i.e. not wild). An important question to ask is: can I change the scale of my data to have it look more normal. Ex: log of the data instead of the data A generic transformation used in statistics is the Box-Cox transformation: Continuous in s s=0 : log s=-1: 1/x s=1: identity

46 Prediction Intervals for File Transfer Times order statistic mean and standard deviation on rescaled data mean and standard deviation

47 Which Summarization Should I Use ? Two issues Robustness to outliers Compactness

48 QQplot is common tool for verifying assumption Normal Qqplot X-axis: standard normal quantiles Y-axis: Ordered statistic of sample: If data comes from a normal distribution, qqplot is close to a straight line (except for end points) Visual inspection is often enough If not possible or doubtful, we will use tests later

49 QQPlots of File Transfer Times

50 Take Home Message The interpretation of  as measure of variability is meaningful if the data is normal (or close to normal). Else, it is misleading. The data should be best re- scaled.

5. Which Summarization to Use ? Issues Robustness to outliers Distribution assumptions 51

A Distribution with Infinite Variance 52 True mean True median True mean True median CI based on std dv CI based on bootsrp CI for median

53 Outlier in File Transfer Time

54 Robustness of Conf/Prediction Intervals mean + std dev CI for median geom mean Outlier removed Outlier present Order stat Based on mean + std dev Based on mean + std dev + re-scaling

Fairness Indices Confidence Intervals obtained by Bootstrap How ? JFI is very dependent on one outlier As expected, since JFI is essentially CoV, i.e. standard deviation Gap is sensitive, but less Does not use squaring ; why ? 55

56 Compactness If normal assumption (or, for CI; asymptotic regime) holds,  and  are more compact two values give both: CIs at all levels, prediction intervals Derived indices: CoV, JFI In contrast, CIs for median does not give information on variability Prediction interval based on order statistic is robust (and, IMHO, best)

57 Take-Home Message Use methods that you understand Mean and standard deviation make sense when data sets are not wild Close to normal, or not heavy tailed and large data sample Use quantiles and order statistics if you have the choice Rescale

Questions 58

Questions 59

Questions 60

More Questions Q2.2 We are given the results of performance measurements for two protocols in the form of series x_1,…,x_m, y_1,..,y_n. We are given the means and sample standard deviations of the two series. How can I decide which protocol is better ? Q2.3 We have tested a system for errors and found 0 in 37 runs. Give a confidence interval for the probability of error. Same question if we find 1 error in 37 runs. 61