A6523 Modeling, Inference, and Mining Jim Cordes, Cornell University. Motivations: Detection & Characterization. Lecture 2.
|
|
- Benjamin West
- 6 years ago
- Views:
Transcription
1 A6523 Modeling, Inference, and Mining Jim Cordes, Cornell University Lecture 2 Probability basics Fourier transform basics Typical problems Overall mantra: Discovery and cri@cal thinking with data + The more I prac,ce the luckier I get Arnold Palmer 1 Motivations: Detection & Characterization Detec@on problem: We seek the existence of a signal in some data. It may not exist in the data or it may be very weak We want to detect it with maximum robustness if it is there If not, we want a robust descrip@on of how strong it could be and be consistent with non- detec@on Characteriza@on problem: We know that a signal is in the data and we want to determine its proper@es op@mally. What do robust and op@mal mean? How we incorporate prior knowledge into an analysis is a key difference between frequen@st and Bayesian methods 2 1
2 Broad Classes of Problems Many measured quantitites ( raw data ) are the outputs of linear systems Wave propagation (EM, gravitational, seismic, acoustic ) Many signals are the result of nonlinear operations in natural systems or in apparati Many analyses of data are linear operations acting on the data to produce some desired result (detection, modeling) E.g. Fourier transform based spectral analysis Many analyses are nonlinear E.g. Maximum entropy and Bayesian spectral analysis 3 Broad Classes of Problems Detection, analysis and modeling: signal detection analysis Natural or artificial Is it there? What are its properties? Optimal detection schemes Maximize S/N of a test statistic Population of signals: maximize detections of real signals minimize false positives and false negatives null hypothesis: no signal there Parametric approaches: (e.g. least squares fitting of a model with parameters) Non-parametric approaches: (e.g. relative comparison of distributions [KS test]) 4 2
3 Probability Definitions Item Defini*on/property comments Random variable Cumula*ve probability distribu*on F X (x) Probability density func*on f X (x) = df X (x) /dx Event à number X = RV, x = instance CDF = cumula@ve probability PDF, Normalized to unity Events can be quan@ta@ve or qualita@ve Characteriza@on, inference, genera@on of random samples di^o Characteris*c func*on Fourier transform of PDF moment genera@ng func@on Joint/mul*variate PDF, CDF Moments Comparison of data sets Collec@ons of random variables Averages of any quan@ty over a PDF Moments, PDFs, CDFs Stochas@c processes and correla@on func@ons e.g. Kolmogorov- Smirnov test for CDFs 5 Notions of Probability Qualitative: (Knowledge) As a measure of the degree to which propositions, hypotheses, or quantities are known. The measure can be an intrinsic property of an event in and of itself or relative to other events. eg. It is not probable that the sun will explode as a supernova is an absolute statement based on our background knowledge of stellar evolution. eg. Horse A is more probable to win, place, or show than horse B is a relative statement that includes the constraint that one horse wins to the exclusion of any other horse in the race. Bayesian inference makes use of this qualitative form of probability along with quantitative aspects discussed below. In this sense, Bayesian methods attribute probabilities to more entities than do some of the more formalistic approaches
4 Quantitative: (Frequentist approach: random variables and ensembles) Classical: Probabilities of events in an event space S are found, a priori, by considerations of possible outcomes of experiments and the manner in which the outcomes may be achieved. For example, a single-die yields probabilities of 1/6 for the outcomes of a single toss of the die. A problem with the classical definition (which often calculates probabilities from line segment lengths, areas, volumes, etc.) is that they cannot be evaluated for all cases (eg. an unfair die). Relative Frequency: The probability of an event is defined to be the limit of the frequency of occurrence of the event in N repeated experiments as N : P { } = lim N N N. The problem with frequencies is that in real experiments N is always finite and therefore probabilities can only be estimated. Such estimates are a poor basis for deductive probabilistic theories. 2 7 Axiomatic: A deductive theory of probability can be based on axioms of probability for allowed events in an overall event space. The axioms are: For = an element in the event space S and P ( ) = probability of the event the following hold: i) 0 P ( ) 1 ii) P {S} =1(S = set of all events, so P {S} is the probability that any event will be the outcome of an experiment. iii) If 1 and 2 are mutually exclusive then the probability of ( the event that 1 or 1 occurs) is P ( )=P ( 1 )+P ( 2 ). These axioms imply further that: iv) If = event that does not occur, then P ( ) =1 P ( ). v) If 1 is a sufficient condition for 2, then P ( 1 ) P ( 2 ). Equality holds when 2 is also a necessary condition for 2. vi) Let P ( 1 2 ) = probability of the event that 1 and 2 occurs (overlap in Venn diagram). Mutually exclusive events have P ( 1 2 )=0while generally P ( 1 2 ) 0. In general we also have P ( )=P ( 1 )+P ( 2 ) P ( 1 2 ) P ( 1 )+P ( 2 ) 3 8 4
5 9 Conditional Probabilities: Also satisfying the axioms are conditional probabilities. Consider the probability that an event 2 occurs given that the event 1 occurs. It may be shown that this probability is Similarly, Bayes Theorem: P { 2 1 } = P { 1 2} P { 1 }. P { 1 2 } = P { 2 1} P { 2 }. Solving for P { 1 2 } from the two preceding equations and setting these solutions equal yields Bayes theorem P { 2 1 } = P { 1 2 }P { 2 }. P { 1 }
6 Bayesian Inference: Bayesian inference is based on the preceding equation where, with relaxed definitions of the event space, hypotheses and parameters are attributed probabilities based on knowledge before and after an experiment is conducted. Bayes theorem combined with the qualitative interpretation of probability therefore allows the sequential acquisition of knowledge (ie. learning) to be handled. The implied temporal sequence of events, by which data are accumulated and the likelihood of a hypothesis being true increases or decreases, represents the power of the Bayesian outlook. Moreover, with Bayesian inference, assumptions behind the inference are often brought up front as conditions upon which probabilities are calculated. 11 Probability on One Page 1. PDF and CDF: f X (x) has unit area and F X (x) is in [0, 1]; f X (x) = df X(x) dx 2. Characteristic function: X(!) = R dx f X (x)e i!x e i!x d 3. Moments: hx n i = i n n X (!) d! n (moment generating function) 3. Change of variable: y = g(x) with solutions x j = g 1 (y),j =1,...,n. Probability is conserved: f Y (y)dy = f X (x)dx: nx f X (x j ) f Y (y) = dg(x)/dx x=xj!=0 NxN transformation: the derivative is replaced by the determinant of the Jacobian matrix. 4. Sum of independent RVs: Z = X + Y: Easily extendable to a sum of N RVs. 5. Conditional PDFs: f X (x y)f Y (y) =f XY (x, y) 6. Bayes Theorem: j=1 f Z (z) = f X f Y Convolution =) Z (!) = X (!) Y (!) Product =) Z 2 = X 2 + Y 2 Variances add f Y (y x) = f X(x y)f Y (y) f X (x)
7 Generating Pseudo Random Numbers Cf. Section 5.13 of Gregory Several methods: 1. transformation method (exponential example) E.g. X = g(y) 2. CDF mapping of uniformly distributed numbers Works for arbitrary output PDF 3. Measurements of natural systems Higher-order statistics (e.g. specifying a power spectral shape): other methods are needed, to be discussed. 13 Arbitrary output PDF U in [0, 1] 14 7
8 1/31/
9 Basic Probability Tools Random variables, event space: ζ = event à X = random var. PDF, CDF, characteris@c func@on Median, mode, mean Condi@onal probabili@es and PDFs Bayes theorem Comparing PDFs Moments and moment tests Sums of random variables and convolu@on theorem Central Limit Theorem Changes of variable Func@ons of random variables Sequences of random variables Stochas@c processes = sequences of random variables vs. t, f, etc. Power spectrum, autocorrela@on, autocovariance, and structure func@ons Bispectrum = 2D spectrum of third moment Random walks, shot noise, autoregressive, moving average, Markov processes 17 Relevant PDFs Gaussian or Normal: N(μ, σ 2 ) f X (x) = 1 p 2 e (x µ)2 /2 2 Random variable argument Exponen@al: Chi 2 : X = NX j=1 f X (x) = f X (x) =hxi 1 e x/hxi H(x) x 2 j with x j = = i.i.d GRV: N(0,1)! 1 (N/2)2 N/2 x(n 2)/2 e x/2 18 9
10 h^p://en.wikipedia.org/wiki/chi- 19 h^p://upload.wikimedia.org/wikipedia/commons/a/a9/empirical_rule.png Confidence intervals 20 10
11 of a Gaussian or Normal RV From wikipedia 21 χ
12 Discrete RV: Poisson Processes If we throw points onto the time axis randomly but with a uniform average rate of having k points in an interval [0,T] is P (k) = e t ( t) k k! the probability (1) Note this is the limiting case of a binomial distribution P (k) = n p k (1 p) n k k where p is the probability of having one point in the interval [0,t] when n points occur in a larger interval [0,T]; i.e. p = t/t. In the limit where n so that p 0 while np = constant > 0, the Poisson expression follows from the binomial probability. The mean and second moment of k are k = t k 2 = k 2 + t. 23 Central Limit Theorem The Central Limit Theorem is a powerful tool for observational science because it can be used to invoke a priori distributions for measured quantities. In simple terms, any quantity that is the sum of many independent ones will have Gaussian statistics. We need to understand what constitutes independence and how Gaussian statistics play out for stochastic processes, which may be viewed as sequences of a large number of random variables. In addition, the CLT does not always apply! Consider the sum Z N = 1 N N X j j=1 where the X j are independent, identically distributed (iid) random variables with means and variances µ j X j j 2 = Xj 2 X j 2. and the PDFs of the X j s are almost arbitrary. Restrictions on the distributions of each X j are that i) j 2 >m>0 m = constant ii) X n <M= constant for n>2 In the limit N,Z N becomes a Gaussian random variable with mean and variance Z N = 1 N µ j, Z 2 = 1 N j=1 N N j=1 2 j
13 Example of arithmetic sums of uniformly distributed numbers Consider the sum x = 1 N x j N j=1 where the x j are drawn from a uniform PDF: x j [ 1/2, 1/2]. The figure shows counts of x for averages of length N =1, 2, 4, 8 and 16 for 100 realizations Example of arithmetic sums of uniformly distributed numbers (continued) Histograms based on 10 5 realizations are shown in the figure below. Two important features are: 1. the width of the PDF scales as N 1/2 2. the shape of the PDF tends toward a Gaussian form with x =0and x = x 2 1/2 1/ 12N
14 27 Cauchy random numbers A Cauchy distribution is Its characteristic function is Using uniformly distributed numbers u [ transformation f X(x) = X( ) =e x 2. 1/2, 1/2], Cauchy random numbers x can be generated using the x =tan( u). Check: Using the change of variable theorem, where has been used. f X(x) = f du u(u(x) dx 1 = 1+x 2, dx du = d du tan u = 1 cos 2 u =1+x
15 Example of arithmetic sums of Cauchy distributed numbers Histograms based on 10 5 realizations are shown in the figure below. Here neither the width nor the shame of the PDF changes as N gets larger! What s happening? 5 29 First let s understand the case of summing uniformly distributed numbers Again consider X j that are all uniformly distributed between ± 1 2. The PDF of a single random variable is f X(x) = and it is a Fourier transform pair with its characteristic function or Graphically: (x) j( ) = e i sin /2 xj = /2 f X(x) = (x) sin f f = sin 2 /
16 Now consider sums of N RVs of this type (i.e. X j) and use the convolution theorem to evaluate the character- j istic function of the sum: Graphically: Gaussian N =2 N =3 N = e x2 ( sin 2 /2 )2 sin /2 ( /2 )3 e We need to rescale the sum to find the characteristic function of Z N = 1 N x j N j=1 defined earlier. From the convolution results we have sin /2 N ( ) = NZN /2 From the transformation of random variables we have that and by the scaling theorem for Fourier transforms f ZN (x) = Nf NZN ( Nx) ZN ( ) = NZN = N sin /2 N N. /2 N
17 If the CLT holds: Now or lim N ZN( ) =e f ZN (x) = e x2 /2 Z Z 2 Z Consistency with this limiting form can be seen by expanding ZN for small ZN ( ) /2 N 1 3! ( /2 N) 3 /2 N 2 1 N 24 that is identical to the expansion of exp ( 2 2 Z/2) Why the CLT does not work for a sum of Cauchy variables: The Cauchy distribution and its characteristic function are f X(x) = x 2 (w) = e In this case has a characteristic function Z N = 1 N x j N j=1 N( ) =e N / N By inspection the exponential will not converge to a Gaussian. Instead, the sum of N Cauchy RVs is a Cauchy RV. Is the Cauchy distribution a legitimate PDF? No! The variance diverges: X 2 = dx x x 2. The Cauchy distribution is an example of a stable distribution, defined as one that has the property that a linear combination of two independent copies of the variable has the same distribution to within a location and a scale parameter. The family of stable distributions is sometimes called the Levy alpha-stable distribution. (Levy flights, etc.) There is a generalized CLT involving PDFs with long power- law tails h^p://en.wikipedia.org/wiki/lévy_distribu@on 10 h^p://en.wikipedia.org/wiki/stable_distribu@on 34 17
18 I. Ensemble vs. Time Averages we are forced to do sample averages of various types Our goal is onen, however, to learn about the parent or ensemble from which the data are conceptually drawn In some averages converge to good of ensemble averages In others, convergence can be very slow or can fail (e.g. red- noise processes) 35 I(t, ζ) 36 18
19 I(t, ζ) 37 I(t, ζ) As data span length T à time average à ensemble average Ergodic 38 19
20 1/31/17 Types of Random Processes Goodman Sta,s,cal Op,cs 39 Example: the Universe Measurements of the CMB and large- scale structure are on a single realiza@on The goal of cosmology is to learn about the (no@onal) ensemble of condi@ons that lead to what we see Quan@ta@vely these are cast in ques@ons like what was the primordial spectrum of density fluctua@ons? and that spectrum is usually parameterized as a power law Perhaps the mul@verse = the ensemble Are all universes the same (sta@s@cally)? Do measurements on our universe typify all universes? (Conven@onal wisdom says no) 40 20
Probability, CLT, CLT counterexamples, Bayes. The PDF file of this lecture contains a full reference document on probability and random variables.
Lecture 5 A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2015 http://www.astro.cornell.edu/~cordes/a6523 Probability, CLT, CLT counterexamples, Bayes The PDF file of
More informationA6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011
A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011 Reading Chapter 5 (continued) Lecture 8 Key points in probability CLT CLT examples Prior vs Likelihood Box & Tiao
More informationA6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011
A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011 Reading Chapter 5 of Gregory (Frequentist Statistical Inference) Lecture 7 Examples of FT applications Simulating
More informationA6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring
A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2015 http://www.astro.cornell.edu/~cordes/a6523 Lecture 4 See web page later tomorrow Searching for Monochromatic Signals
More informationDeep Learning for Computer Vision
Deep Learning for Computer Vision Lecture 3: Probability, Bayes Theorem, and Bayes Classification Peter Belhumeur Computer Science Columbia University Probability Should you play this game? Game: A fair
More informationA6523 Modeling, Inference, and Mining Jim Cordes, Cornell University. False Positives in Fourier Spectra. For N = DFT length: Lecture 5 Reading
A6523 Modeling, Inference, and Mining Jim Cordes, Cornell University Lecture 5 Reading Notes on web page Stochas
More informationReview. DS GA 1002 Statistical and Mathematical Models. Carlos Fernandez-Granda
Review DS GA 1002 Statistical and Mathematical Models http://www.cims.nyu.edu/~cfgranda/pages/dsga1002_fall16 Carlos Fernandez-Granda Probability and statistics Probability: Framework for dealing with
More informationMultiple Random Variables
Multiple Random Variables Joint Probability Density Let X and Y be two random variables. Their joint distribution function is F ( XY x, y) P X x Y y. F XY ( ) 1, < x
More informationELEG 3143 Probability & Stochastic Process Ch. 6 Stochastic Process
Department of Electrical Engineering University of Arkansas ELEG 3143 Probability & Stochastic Process Ch. 6 Stochastic Process Dr. Jingxian Wu wuj@uark.edu OUTLINE 2 Definition of stochastic process (random
More informationA523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011
A523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2011 Lecture 1 Organization:» Syllabus (text, requirements, topics)» Course approach (goals, themes) Book: Gregory, Bayesian
More informationconditional cdf, conditional pdf, total probability theorem?
6 Multiple Random Variables 6.0 INTRODUCTION scalar vs. random variable cdf, pdf transformation of a random variable conditional cdf, conditional pdf, total probability theorem expectation of a random
More informationPreliminary statistics
1 Preliminary statistics The solution of a geophysical inverse problem can be obtained by a combination of information from observed data, the theoretical relation between data and earth parameters (models),
More informationCourse: ESO-209 Home Work: 1 Instructor: Debasis Kundu
Home Work: 1 1. Describe the sample space when a coin is tossed (a) once, (b) three times, (c) n times, (d) an infinite number of times. 2. A coin is tossed until for the first time the same result appear
More informationA6523 Modeling, Inference, and Mining Jim Cordes, Cornell University
A6523 Modeling, Inference, and Mining Jim Cordes, Cornell University Lecture 19 Modeling Topics plan: Modeling (linear/non- linear least squares) Bayesian inference Bayesian approaches to spectral esbmabon;
More informationBayesian Methods for Machine Learning
Bayesian Methods for Machine Learning CS 584: Big Data Analytics Material adapted from Radford Neal s tutorial (http://ftp.cs.utoronto.ca/pub/radford/bayes-tut.pdf), Zoubin Ghahramni (http://hunch.net/~coms-4771/zoubin_ghahramani_bayesian_learning.pdf),
More informationPart IA Probability. Definitions. Based on lectures by R. Weber Notes taken by Dexter Chua. Lent 2015
Part IA Probability Definitions Based on lectures by R. Weber Notes taken by Dexter Chua Lent 2015 These notes are not endorsed by the lecturers, and I have modified them (often significantly) after lectures.
More informationSTATS 200: Introduction to Statistical Inference. Lecture 29: Course review
STATS 200: Introduction to Statistical Inference Lecture 29: Course review Course review We started in Lecture 1 with a fundamental assumption: Data is a realization of a random process. The goal throughout
More informationOrder Statistics and Distributions
Order Statistics and Distributions 1 Some Preliminary Comments and Ideas In this section we consider a random sample X 1, X 2,..., X n common continuous distribution function F and probability density
More informationIf we want to analyze experimental or simulated data we might encounter the following tasks:
Chapter 1 Introduction If we want to analyze experimental or simulated data we might encounter the following tasks: Characterization of the source of the signal and diagnosis Studying dependencies Prediction
More informationPart IA Probability. Theorems. Based on lectures by R. Weber Notes taken by Dexter Chua. Lent 2015
Part IA Probability Theorems Based on lectures by R. Weber Notes taken by Dexter Chua Lent 2015 These notes are not endorsed by the lecturers, and I have modified them (often significantly) after lectures.
More informationProbability and statistics; Rehearsal for pattern recognition
Probability and statistics; Rehearsal for pattern recognition Václav Hlaváč Czech Technical University in Prague Czech Institute of Informatics, Robotics and Cybernetics 166 36 Prague 6, Jugoslávských
More informationMath 494: Mathematical Statistics
Math 494: Mathematical Statistics Instructor: Jimin Ding jmding@wustl.edu Department of Mathematics Washington University in St. Louis Class materials are available on course website (www.math.wustl.edu/
More informationAsymptotic Statistics-VI. Changliang Zou
Asymptotic Statistics-VI Changliang Zou Kolmogorov-Smirnov distance Example (Kolmogorov-Smirnov confidence intervals) We know given α (0, 1), there is a well-defined d = d α,n such that, for any continuous
More informationCS 6140: Machine Learning Spring 2016
CS 6140: Machine Learning Spring 2016 Instructor: Lu Wang College of Computer and Informa?on Science Northeastern University Webpage: www.ccs.neu.edu/home/luwang Email: luwang@ccs.neu.edu Logis?cs Assignment
More informationMonte Carlo Simulation. CWR 6536 Stochastic Subsurface Hydrology
Monte Carlo Simulation CWR 6536 Stochastic Subsurface Hydrology Steps in Monte Carlo Simulation Create input sample space with known distribution, e.g. ensemble of all possible combinations of v, D, q,
More informationHypothesis Testing. 1 Definitions of test statistics. CB: chapter 8; section 10.3
Hypothesis Testing CB: chapter 8; section 0.3 Hypothesis: statement about an unknown population parameter Examples: The average age of males in Sweden is 7. (statement about population mean) The lowest
More informationStatistics 100A Homework 5 Solutions
Chapter 5 Statistics 1A Homework 5 Solutions Ryan Rosario 1. Let X be a random variable with probability density function a What is the value of c? fx { c1 x 1 < x < 1 otherwise We know that for fx to
More informationIntroduction to Systems Analysis and Decision Making Prepared by: Jakub Tomczak
Introduction to Systems Analysis and Decision Making Prepared by: Jakub Tomczak 1 Introduction. Random variables During the course we are interested in reasoning about considered phenomenon. In other words,
More informationPreliminary Statistics Lecture 2: Probability Theory (Outline) prelimsoas.webs.com
1 School of Oriental and African Studies September 2015 Department of Economics Preliminary Statistics Lecture 2: Probability Theory (Outline) prelimsoas.webs.com Gujarati D. Basic Econometrics, Appendix
More informationE X A M. Probability Theory and Stochastic Processes Date: December 13, 2016 Duration: 4 hours. Number of pages incl.
E X A M Course code: Course name: Number of pages incl. front page: 6 MA430-G Probability Theory and Stochastic Processes Date: December 13, 2016 Duration: 4 hours Resources allowed: Notes: Pocket calculator,
More informationA6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring
Lecture 8 A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 2015 http://www.astro.cornell.edu/~cordes/a6523 Applications: Bayesian inference: overview and examples Introduction
More informationH 2 : otherwise. that is simply the proportion of the sample points below level x. For any fixed point x the law of large numbers gives that
Lecture 28 28.1 Kolmogorov-Smirnov test. Suppose that we have an i.i.d. sample X 1,..., X n with some unknown distribution and we would like to test the hypothesis that is equal to a particular distribution
More informationNorthwestern University Department of Electrical Engineering and Computer Science
Northwestern University Department of Electrical Engineering and Computer Science EECS 454: Modeling and Analysis of Communication Networks Spring 2008 Probability Review As discussed in Lecture 1, probability
More informationIntroduction to Probability and Statistics (Continued)
Introduction to Probability and Statistics (Continued) Prof. icholas Zabaras Center for Informatics and Computational Science https://cics.nd.edu/ University of otre Dame otre Dame, Indiana, USA Email:
More informationSTAT2201. Analysis of Engineering & Scientific Data. Unit 3
STAT2201 Analysis of Engineering & Scientific Data Unit 3 Slava Vaisman The University of Queensland School of Mathematics and Physics What we learned in Unit 2 (1) We defined a sample space of a random
More information8 Laws of large numbers
8 Laws of large numbers 8.1 Introduction We first start with the idea of standardizing a random variable. Let X be a random variable with mean µ and variance σ 2. Then Z = (X µ)/σ will be a random variable
More informationCS 6140: Machine Learning Spring What We Learned Last Week. Survey 2/26/16. VS. Model
Logis@cs CS 6140: Machine Learning Spring 2016 Instructor: Lu Wang College of Computer and Informa@on Science Northeastern University Webpage: www.ccs.neu.edu/home/luwang Email: luwang@ccs.neu.edu Assignment
More informationRecitation 2: Probability
Recitation 2: Probability Colin White, Kenny Marino January 23, 2018 Outline Facts about sets Definitions and facts about probability Random Variables and Joint Distributions Characteristics of distributions
More informationChapter 6: Random Processes 1
Chapter 6: Random Processes 1 Yunghsiang S. Han Graduate Institute of Communication Engineering, National Taipei University Taiwan E-mail: yshan@mail.ntpu.edu.tw 1 Modified from the lecture notes by Prof.
More informationReview of probability
Review of probability Computer Sciences 760 Spring 2014 http://pages.cs.wisc.edu/~dpage/cs760/ Goals for the lecture you should understand the following concepts definition of probability random variables
More informationProbability Models in Electrical and Computer Engineering Mathematical models as tools in analysis and design Deterministic models Probability models
Probability Models in Electrical and Computer Engineering Mathematical models as tools in analysis and design Deterministic models Probability models Statistical regularity Properties of relative frequency
More informationRandom Variables and Their Distributions
Chapter 3 Random Variables and Their Distributions A random variable (r.v.) is a function that assigns one and only one numerical value to each simple event in an experiment. We will denote r.vs by capital
More information1 Presessional Probability
1 Presessional Probability Probability theory is essential for the development of mathematical models in finance, because of the randomness nature of price fluctuations in the markets. This presessional
More information6.1 Moment Generating and Characteristic Functions
Chapter 6 Limit Theorems The power statistics can mostly be seen when there is a large collection of data points and we are interested in understanding the macro state of the system, e.g., the average,
More informationLecture 2: Repetition of probability theory and statistics
Algorithms for Uncertainty Quantification SS8, IN2345 Tobias Neckel Scientific Computing in Computer Science TUM Lecture 2: Repetition of probability theory and statistics Concept of Building Block: Prerequisites:
More informationStatistics for scientists and engineers
Statistics for scientists and engineers February 0, 006 Contents Introduction. Motivation - why study statistics?................................... Examples..................................................3
More informationPractice Problems Section Problems
Practice Problems Section 4-4-3 4-4 4-5 4-6 4-7 4-8 4-10 Supplemental Problems 4-1 to 4-9 4-13, 14, 15, 17, 19, 0 4-3, 34, 36, 38 4-47, 49, 5, 54, 55 4-59, 60, 63 4-66, 68, 69, 70, 74 4-79, 81, 84 4-85,
More informationRecall the Basics of Hypothesis Testing
Recall the Basics of Hypothesis Testing The level of significance α, (size of test) is defined as the probability of X falling in w (rejecting H 0 ) when H 0 is true: P(X w H 0 ) = α. H 0 TRUE H 1 TRUE
More informationIntroduction to Bayesian Learning. Machine Learning Fall 2018
Introduction to Bayesian Learning Machine Learning Fall 2018 1 What we have seen so far What does it mean to learn? Mistake-driven learning Learning by counting (and bounding) number of mistakes PAC learnability
More informationMcGill University. Faculty of Science. Department of Mathematics and Statistics. Part A Examination. Statistics: Theory Paper
McGill University Faculty of Science Department of Mathematics and Statistics Part A Examination Statistics: Theory Paper Date: 10th May 2015 Instructions Time: 1pm-5pm Answer only two questions from Section
More informationF & B Approaches to a simple model
A6523 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Spring 215 http://www.astro.cornell.edu/~cordes/a6523 Lecture 11 Applications: Model comparison Challenges in large-scale surveys
More informationSample Spaces, Random Variables
Sample Spaces, Random Variables Moulinath Banerjee University of Michigan August 3, 22 Probabilities In talking about probabilities, the fundamental object is Ω, the sample space. (elements) in Ω are denoted
More informationSome Review and Hypothesis Tes4ng. Friday, March 15, 13
Some Review and Hypothesis Tes4ng Outline Discussing the homework ques4ons from Joey and Phoebe Review of Sta4s4cal Inference Proper4es of OLS under the normality assump4on Confidence Intervals, T test,
More informationSystem Identification
System Identification Arun K. Tangirala Department of Chemical Engineering IIT Madras July 27, 2013 Module 3 Lecture 1 Arun K. Tangirala System Identification July 27, 2013 1 Objectives of this Module
More informationSTATISTICS SYLLABUS UNIT I
STATISTICS SYLLABUS UNIT I (Probability Theory) Definition Classical and axiomatic approaches.laws of total and compound probability, conditional probability, Bayes Theorem. Random variable and its distribution
More informationLecture 1: Probability Fundamentals
Lecture 1: Probability Fundamentals IB Paper 7: Probability and Statistics Carl Edward Rasmussen Department of Engineering, University of Cambridge January 22nd, 2008 Rasmussen (CUED) Lecture 1: Probability
More informationIntelligent Embedded Systems Uncertainty, Information and Learning Mechanisms (Part 1)
Advanced Research Intelligent Embedded Systems Uncertainty, Information and Learning Mechanisms (Part 1) Intelligence for Embedded Systems Ph. D. and Master Course Manuel Roveri Politecnico di Milano,
More information* Tuesday 17 January :30-16:30 (2 hours) Recored on ESSE3 General introduction to the course.
Name of the course Statistical methods and data analysis Audience The course is intended for students of the first or second year of the Graduate School in Materials Engineering. The aim of the course
More informationLearning Objectives for Stat 225
Learning Objectives for Stat 225 08/20/12 Introduction to Probability: Get some general ideas about probability, and learn how to use sample space to compute the probability of a specific event. Set Theory:
More informationStat 5101 Lecture Notes
Stat 5101 Lecture Notes Charles J. Geyer Copyright 1998, 1999, 2000, 2001 by Charles J. Geyer May 7, 2001 ii Stat 5101 (Geyer) Course Notes Contents 1 Random Variables and Change of Variables 1 1.1 Random
More informationMath 152. Rumbos Fall Solutions to Exam #2
Math 152. Rumbos Fall 2009 1 Solutions to Exam #2 1. Define the following terms: (a) Significance level of a hypothesis test. Answer: The significance level, α, of a hypothesis test is the largest probability
More informationSTAT 4385 Topic 01: Introduction & Review
STAT 4385 Topic 01: Introduction & Review Xiaogang Su, Ph.D. Department of Mathematical Science University of Texas at El Paso xsu@utep.edu Spring, 2016 Outline Welcome What is Regression Analysis? Basics
More informationCS145: Probability & Computing
CS45: Probability & Computing Lecture 5: Concentration Inequalities, Law of Large Numbers, Central Limit Theorem Instructor: Eli Upfal Brown University Computer Science Figure credits: Bertsekas & Tsitsiklis,
More informationS n = x + X 1 + X X n.
0 Lecture 0 0. Gambler Ruin Problem Let X be a payoff if a coin toss game such that P(X = ) = P(X = ) = /2. Suppose you start with x dollars and play the game n times. Let X,X 2,...,X n be payoffs in each
More informationStatistical Methods in Particle Physics
Statistical Methods in Particle Physics. Probability Distributions Prof. Dr. Klaus Reygers (lectures) Dr. Sebastian Neubert (tutorials) Heidelberg University WS 07/8 Gaussian g(x; µ, )= p exp (x µ) https://en.wikipedia.org/wiki/normal_distribution
More informationwhere r n = dn+1 x(t)
Random Variables Overview Probability Random variables Transforms of pdfs Moments and cumulants Useful distributions Random vectors Linear transformations of random vectors The multivariate normal distribution
More informationFourier and Stats / Astro Stats and Measurement : Stats Notes
Fourier and Stats / Astro Stats and Measurement : Stats Notes Andy Lawrence, University of Edinburgh Autumn 2013 1 Probabilities, distributions, and errors Laplace once said Probability theory is nothing
More informationfor valid PSD. PART B (Answer all five units, 5 X 10 = 50 Marks) UNIT I
Code: 15A04304 R15 B.Tech II Year I Semester (R15) Regular Examinations November/December 016 PROBABILITY THEY & STOCHASTIC PROCESSES (Electronics and Communication Engineering) Time: 3 hours Max. Marks:
More informationLecture 1: August 28
36-705: Intermediate Statistics Fall 2017 Lecturer: Siva Balakrishnan Lecture 1: August 28 Our broad goal for the first few lectures is to try to understand the behaviour of sums of independent random
More informationChapter 2 Random Variables
Stochastic Processes Chapter 2 Random Variables Prof. Jernan Juang Dept. of Engineering Science National Cheng Kung University Prof. Chun-Hung Liu Dept. of Electrical and Computer Eng. National Chiao Tung
More informationPerhaps the simplest way of modeling two (discrete) random variables is by means of a joint PMF, defined as follows.
Chapter 5 Two Random Variables In a practical engineering problem, there is almost always causal relationship between different events. Some relationships are determined by physical laws, e.g., voltage
More informationProbability and Distributions
Probability and Distributions What is a statistical model? A statistical model is a set of assumptions by which the hypothetical population distribution of data is inferred. It is typically postulated
More informationECE353: Probability and Random Processes. Lecture 7 -Continuous Random Variable
ECE353: Probability and Random Processes Lecture 7 -Continuous Random Variable Xiao Fu School of Electrical Engineering and Computer Science Oregon State University E-mail: xiao.fu@oregonstate.edu Continuous
More informationIrr. Statistical Methods in Experimental Physics. 2nd Edition. Frederick James. World Scientific. CERN, Switzerland
Frederick James CERN, Switzerland Statistical Methods in Experimental Physics 2nd Edition r i Irr 1- r ri Ibn World Scientific NEW JERSEY LONDON SINGAPORE BEIJING SHANGHAI HONG KONG TAIPEI CHENNAI CONTENTS
More informationPractical Statistics
Practical Statistics Lecture 1 (Nov. 9): - Correlation - Hypothesis Testing Lecture 2 (Nov. 16): - Error Estimation - Bayesian Analysis - Rejecting Outliers Lecture 3 (Nov. 18) - Monte Carlo Modeling -
More informationAlgorithms for Uncertainty Quantification
Algorithms for Uncertainty Quantification Tobias Neckel, Ionuț-Gabriel Farcaș Lehrstuhl Informatik V Summer Semester 2017 Lecture 2: Repetition of probability theory and statistics Example: coin flip Example
More informationSignal Modeling, Statistical Inference and Data Mining in Astrophysics
ASTRONOMY 6523 Spring 2013 Signal Modeling, Statistical Inference and Data Mining in Astrophysics Course Approach The philosophy of the course reflects that of the instructor, who takes a dualistic view
More informationInstitute of Actuaries of India
Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics For 2018 Examinations Subject CT3 Probability and Mathematical Statistics Core Technical Syllabus 1 June 2017 Aim The
More informationPattern Recognition and Machine Learning. Bishop Chapter 2: Probability Distributions
Pattern Recognition and Machine Learning Chapter 2: Probability Distributions Cécile Amblard Alex Kläser Jakob Verbeek October 11, 27 Probability Distributions: General Density Estimation: given a finite
More informationChapter 9. Non-Parametric Density Function Estimation
9-1 Density Estimation Version 1.2 Chapter 9 Non-Parametric Density Function Estimation 9.1. Introduction We have discussed several estimation techniques: method of moments, maximum likelihood, and least
More information1 Introduction. P (n = 1 red ball drawn) =
Introduction Exercises and outline solutions. Y has a pack of 4 cards (Ace and Queen of clubs, Ace and Queen of Hearts) from which he deals a random of selection 2 to player X. What is the probability
More informationData Analysis I. Dr Martin Hendry, Dept of Physics and Astronomy University of Glasgow, UK. 10 lectures, beginning October 2006
Astronomical p( y x, I) p( x, I) p ( x y, I) = p( y, I) Data Analysis I Dr Martin Hendry, Dept of Physics and Astronomy University of Glasgow, UK 10 lectures, beginning October 2006 4. Monte Carlo Methods
More informationConvergence of Random Processes
Convergence of Random Processes DS GA 1002 Probability and Statistics for Data Science http://www.cims.nyu.edu/~cfgranda/pages/dsga1002_fall17 Carlos Fernandez-Granda Aim Define convergence for random
More informationCOMBINING VARIABLES ERROR PROPAGATION. What is the error on a quantity that is a function of several random variables
COMBINING VARIABLES ERROR PROPAGATION What is the error on a quantity that is a function of several random variables θ = f(x, y,...) If the variance on x, y,... is small and uncorrelated variables then
More informationAST 418/518 Instrumentation and Statistics
AST 418/518 Instrumentation and Statistics Class Website: http://ircamera.as.arizona.edu/astr_518 Class Texts: Practical Statistics for Astronomers, J.V. Wall, and C.R. Jenkins Measuring the Universe,
More informationData Processing Techniques
Universitas Gadjah Mada Department of Civil and Environmental Engineering Master in Engineering in Natural Disaster Management Data Processing Techniques Hypothesis Tes,ng 1 Hypothesis Testing Mathema,cal
More informationApplied Probability and Stochastic Processes
Applied Probability and Stochastic Processes In Engineering and Physical Sciences MICHEL K. OCHI University of Florida A Wiley-Interscience Publication JOHN WILEY & SONS New York - Chichester Brisbane
More informationExercises and Answers to Chapter 1
Exercises and Answers to Chapter The continuous type of random variable X has the following density function: a x, if < x < a, f (x), otherwise. Answer the following questions. () Find a. () Obtain mean
More informationReview of Probabilities and Basic Statistics
Alex Smola Barnabas Poczos TA: Ina Fiterau 4 th year PhD student MLD Review of Probabilities and Basic Statistics 10-701 Recitations 1/25/2013 Recitation 1: Statistics Intro 1 Overview Introduction to
More informationConfidence Intervals, Testing and ANOVA Summary
Confidence Intervals, Testing and ANOVA Summary 1 One Sample Tests 1.1 One Sample z test: Mean (σ known) Let X 1,, X n a r.s. from N(µ, σ) or n > 30. Let The test statistic is H 0 : µ = µ 0. z = x µ 0
More informationStatistical techniques for data analysis in Cosmology
Statistical techniques for data analysis in Cosmology arxiv:0712.3028; arxiv:0911.3105 Numerical recipes (the bible ) Licia Verde ICREA & ICC UB-IEEC http://icc.ub.edu/~liciaverde outline Lecture 1: Introduction
More informationReview (probability, linear algebra) CE-717 : Machine Learning Sharif University of Technology
Review (probability, linear algebra) CE-717 : Machine Learning Sharif University of Technology M. Soleymani Fall 2012 Some slides have been adopted from Prof. H.R. Rabiee s and also Prof. R. Gutierrez-Osuna
More informationFundamentals of Applied Probability and Random Processes
Fundamentals of Applied Probability and Random Processes,nd 2 na Edition Oliver C. Ibe University of Massachusetts, LoweLL, Massachusetts ip^ W >!^ AMSTERDAM BOSTON HEIDELBERG LONDON NEW YORK OXFORD PARIS
More informationPART I INTRODUCTION The meaning of probability Basic definitions for frequentist statistics and Bayesian inference Bayesian inference Combinatorics
Table of Preface page xi PART I INTRODUCTION 1 1 The meaning of probability 3 1.1 Classical definition of probability 3 1.2 Statistical definition of probability 9 1.3 Bayesian understanding of probability
More informationFundamental Tools - Probability Theory IV
Fundamental Tools - Probability Theory IV MSc Financial Mathematics The University of Warwick October 1, 2015 MSc Financial Mathematics Fundamental Tools - Probability Theory IV 1 / 14 Model-independent
More informationUniversität Potsdam Institut für Informatik Lehrstuhl Maschinelles Lernen. Bayesian Learning. Tobias Scheffer, Niels Landwehr
Universität Potsdam Institut für Informatik Lehrstuhl Maschinelles Lernen Bayesian Learning Tobias Scheffer, Niels Landwehr Remember: Normal Distribution Distribution over x. Density function with parameters
More informationSTAT Chapter 5 Continuous Distributions
STAT 270 - Chapter 5 Continuous Distributions June 27, 2012 Shirin Golchi () STAT270 June 27, 2012 1 / 59 Continuous rv s Definition: X is a continuous rv if it takes values in an interval, i.e., range
More informationMonte Carlo Integration I [RC] Chapter 3
Aula 3. Monte Carlo Integration I 0 Monte Carlo Integration I [RC] Chapter 3 Anatoli Iambartsev IME-USP Aula 3. Monte Carlo Integration I 1 There is no exact definition of the Monte Carlo methods. In the
More informationDeterministic. Deterministic data are those can be described by an explicit mathematical relationship
Random data Deterministic Deterministic data are those can be described by an explicit mathematical relationship Deterministic x(t) =X cos r! k m t Non deterministic There is no way to predict an exact
More informationStatistical Methods for Astronomy
Statistical Methods for Astronomy Probability (Lecture 1) Statistics (Lecture 2) Why do we need statistics? Useful Statistics Definitions Error Analysis Probability distributions Error Propagation Binomial
More information