Probability Theory  

In General > s.a. integration; measure theory; random processes; statistics.
* Idea: The branch of mathematics that deals with computing the odds of an event's occurrence.
* History: The idea that rigorous math can be used to calculate odds was introduced by Pascal, who conferred with Fermat; De Moivre found the Bell curve for distributions of random variables; The theory was founded by Laplace.
* Applications: Telecommunications (e.g., Dublin IAS applied probability group).
$ Definition (measure theory): A probability is a normalized measure on a set of possible events Γ.
$ Definition (logic): A probability is a map P : X × X → [0,1], where X is a set of propositions, closed under ∧, \(\lor\), and ¬, such that (1) P(bc|a) = P(b|a) P(c|ab); (2) P(b|a) + Pb|a) = 1; (3) Pa|a) = 0.
> Online resources: see Wikipedia page; H. Pishro-Nik's text.

Interpretations and Applications > s.a. probability in physics [including axiomatic approach].
* Bayesian approach: An approach to the problem of inferring something about a parameter or state of nature s after observing a random variable x whose distribution p depends on s; Probabilities are "degrees of belief," and refer to our confidence in certain statements based on previous experience; Useful for measurements and updating our predictions, allows us to assign probabilities that numbers be "true values" and to use induction; & Bayes, Bernoulli, Gauss, Laplace (used it to conclude that the boy-girl ratio [< 1] is universal to humankind and determined by biology); XX-century statistics was overwhelmingly behavioristic and frequentist, especially in applications, but the XXI century is seeing a resurgence of Bayesianism; > s.a. foundations of quantum mechanics; statistics.
* Frequentist approach: Probabilities are frequencies of occurrence of values in the ensembles of all observations; It is considered the conventional one in physics, but if strictly followed will lead nowhere.
* Lindley paradox: A counterintuitive statistical effect where the Bayesian and frequentist approaches to hypothesis testing give radically different answers, depending on the choice of the prior distribution.
* Combinatorial approach: Probabilities are ratios of favorable cases to some statement over all cases, in a series of performed tests.
@ Frequentist: Brody in(89); Wall SHPSA(06) [Venn's opposition to degree of belief].
@ Bayesian: Hartigan 83; D'Agostini hp/95 [primer]; Howson BJPS(97) [and logic]; Weintraub BJPS(01) [paradox]; D'Agostini 03 [primer]; Jeffrey 04 [readable]; Appleby O&S(05)qp/04-in [vs frequentist]; Howson & Urbach 05; D'Agostini ISBA-phy/05 ["Fermi's Bayes theorem"]; Mielczarek et al a0901 [introduction, conceptual]; Fuchs & Schack AIP(09)-a0906 [priors in quantum Bayesian inference]; Henderson et al PhSc(10)apr [Bayesian inference in a hierarchical hypothesis space]; Leitgeb & Pettigrew PhSc(10)apr, PhSc(10)apr [justification]; Alamino a1008 [as a framework for physical theories]; Efron BAMS(13) [250th anniversary]; Benétreau-Dupin Syn(15)-a1412 [adequately handling ignorance]; Mauri & Paris PLA(15)-a1510 [Lindley paradox in optical interferometry]; Lucy A&A(16)-a1511 [frequentist tests for Bayesian models].

Related Concepts
* Marginal distribution: For a subset of a collection of random variables, it is the probability distribution of the variables in the subset, obtained adding the values of the overall probability distribution for all values of the remaining variables; > s.a. Wikipedia page.
* Moments of a distribution: The n-th moment of a probability distribution for x is the mean value of xn.
@ Moments of a distribution: Yehia JPA(02)-a1308 [relation between the first and second moments].
@ Combining probability distributions: Eliazar & Klafter PhyA(08) [limit laws and non-linear scaling schemes]; > s.a. Central-Limit Theorem.
@ Space of probability distributions: Calvo et al JSP(10) [renormalization group transformation]; Pistone Ent(15)-a1308 [application to Statistical Physics].
@ Marginal distribution: Cheon JKPS(15)-a1507 [in quantum mechanics, graphical representation].
> Related topics: see analysis [fractional moments]; Law of Large Numbers; measure theory.

Examples of Probability Distributions > see Binomial, Gamma, gaussian, Poisson Distribution; Exponential Family.

References > s.a. probabilistic combinatorics; probability in quantum theory [quasi-probability].
@ General: Bernoulli 1713; Howson BJPS(95) [survey]; Williams SHPSA(05) [history, Cardano]; Novak & LaCroix a1205-proc [free probability, lectures].
@ Books, I: Ekeland 93; Nahin 00; Holland 02; Prakasa Rao 08; Mukhopadhyay 11; Woolfson 12.
@ Books, II: Grimmett & Welsh 86; Rotar 98; Stirzaker 99, 99 [II]; Dokuchaev 15 [1-semester course].
@ Books: Marle 74; de Groot 75; Bitsakis & Nicolaides ed-89; Fristedt & Grey 97; Grimmett & Stirzaker 01 [and random processes], 01 [exercises]; Rosenthal 06 [rigorous]; Heyer 09 [structural, algebraic-topological aspects].
@ Books, problems: Chaumont & Yor 12 [r CP(13)]; Mills 13.
@ Books, special emphasis: Billingsley 68 [convergence]; Mackey 78 [and group representations]; Stroock 93 [analytic].
@ Results: Tribelsky PRL(02)m.PR/01 [sums of variables].
@ Fuzzy: in Gudder IJTP(00), Gudder FP(00) [rev]; Habil & Nasr IJTP(02).
@ Complex probabilities: Salcedo JMP(97); Weingarten PRL(02) [and path integrals]; Bender et al AP(10)-a0912 [and quantum mechanics].
@ Related topics: Nelson 87 [non-standard]; Friedman AAM(99) [frequency interpretation]; Streater JMP(00)mp [historical survey].

Generalizations > s.a. homotopy [homotopy probability theory]; probability in physics [negative probabilities, quantum probabilities]; Plausibility Measures.
@ General references: Liu a1705 [infinite-dimensional probability, intro].
@ Extended probabilities: Niestegge JPA(01)-a1001 [non-Boolean extension, and quantum measurement]; Burgin a0912 [mathematical foundations]; Noldus a1509 [two physical interpretations].


main pageabbreviationsjournalscommentsother sitesacknowledgements
send feedback and suggestions to bombelli at olemiss.edu – modified 11 may 2017