arXiv subjects
Publications and source records attributed to Doron Zeilberger.
We use three kinds of computations: simulation, numeric, and symbolic, to guide risk-averse gamblers in general, and offer particular advice on how to resolve the famous St. Petersburg paradox.
Using experimental mathematics and symbolic computation, we derive many moments for the duration of a three player (fair) gambler's ruin.
In this {\it case study}, we hope to show why Sheldon Axler was not just wrong, but {\em wrong}, when he urged, in 1995: ``Down with Determinants!''. We first recall how determinants are useful in enumerative combinatorics, and then illustrate three versatile tools (Dodgson's condensation, the holonomic ansatz and constant term evaluations) to operate in tandem to prove a certain intriguing determinantal formula conjectured by the first author. We conclude with a postscript describing yet another, much more efficient, method for evaluating determinants: `ask determinant-guru, Christian Krattenthaler', but advise people only to use it as a last resort, since if we would have used this last method right away, we would not have had the fun of doing it all by ourselves.
We demonstrate the power of Experimental Mathematics and Symbolic Computation to study intriguing problems on rational difference equations, studied extensively by Difference Equations giants, Saber Elaydi and Gerry Ladas (and their students and collaborators). In particular we rigorously prove some fascinating conjectures made by Amal Amleh and Gerry Ladas back in 2000. For other conjectures we are content with semi-rigorous proofs. We also extend the work of Emilie Purvine (formerly Hogan) and Doron Zeilberger for rigorously and semi-rigorously proving global asymptotic stability of arbitrary rational difference equations (with positive coefficients), and more generally rational transformations of the positive orthant of $R^k$ into itself
We show the power of Bruno Buchberger's seminal Groebner Basis algorithm, interfaced, seamlessly, with what we call symbolic dynamical programming, to automatically generate algebraic equations satisfied by the generating functions enumerating so-called Generalized Dyck Walks, i.e. 2D walks that start and end on the x-axis, and never dip below it, for an arbitrary set of steps. More impressively, we combine it with calculus (that Maple knows very well!), to automatically compute generating functions for the sum-of-the-areas of these generalized Dyck paths, and even for the sum of any given power of the areas, enabling us to get statistical information about the area under a random generalized Dyck path.
We extract brilliant ideas of Sandi Klavzar, Michel Mollard, and Marko Petkovsek who used them to solve one very specific enumeration problem, namely counting the number of words in the alphabet {0,1} of length n avoiding two consecutive ones, and having exactly k such neighbors, to a much more general setting where one has any (finite) alphabet, and any (finite) set of forbidden subwords. More important, we fully implement it in Maple.
Using Symbolic Computation with Maple, we can discover lots of (rigorously-proved!) facts about Standard Young Tableaux, in particular the distribution of the entries in any specific cell, and the sorting probabilities.
This paper consists of two independent, but related parts. In the first part we show how to use symbolic computation to derive explicit expressions for the first few moments of the number of implicants that a random Boolean function has, or equivalently the number of fixed-dimensional subcubes contained in a random subset of the $n$-dimensional cube. These explicit expressions suggest, but do not prove, that these random variables are always asymptotically normal. The second part presents a full, human-generated proof, of this asymptotic normality, first proved by Urszula Konieczna.
Noga Alon and Yaakov Malinovsky recently studied the following game: you start at 0, and keep rolling a fair standard die, and add the outcomes until the sum happens to be prime. We generalize this in several ways, illustrating the power of symbolic, rather than merely numeric, computation. We conclude with polemics why the beautiful rigorous error estimate of Alon and Malinovsky is only of theoretical interest, explaining why we were content, in our numerous extensions, with non-rigorous, but practically-certain, estimates.
This is a symbolic-computational redux, and extension, of a beautiful paper, by Harry Dym and Eugene Luks, published in 1966 (but written in 1964) about a certain game with balls and cells.
Given positive integers $r$ and $s$, we use inclusion-exclusion, weighted-counting of tilings, and dynamical programming, in order to enumerate, semi-efficiently, the classes of permutations mentioned in the title. In the process we revisit beautiful previous work of Enrique Navarrete, Robert Tauraso, David Robbins (to whose memory this article is dedicated), and John Riordan. We also present two new proofs of John Riordan's recurrence (from 1965) for the sequence enumerating permutations without rising and falling successions (the $r = 1$, $s = 1$ case of the title in the sense of absolute value). The first is fully automatic using the (continuous) Almkvist-Zeilberger algorithm, while the second is purely human-generated via an elegant combinatorial argument. We continue with some open questions and pledge donations to the OEIS in honor of the solvers. We conclude with a postscript describing interesting ideas of Rintaro Matsuo, that we were made aware of after the first version was written, and announce that one of the challenges was met.
We explore both automated and human approaches to the generalized Missionaries and Cannibals problem.
Using data collected from the AMS websites, and the internet in general, we compute the weight-enumerator of the set of 145 AMS fellows who passed away before Aug. 2022, according to the statistics [Age, Number of Publications, Number of Citations]. Using Maple and elementary calculus, we deduced from it the numbers stated in the title, as well as much more detailed and fascinating statistical information. For example, while the correlation coefficient between the number of publications and number of citations is, not surprisingly, positive, it is far lower than one would expect, being a mere 0.5797 . The appendices contain the rankings according to longevity, number of publications, and number of citations.
We show that identities involving trigonometric sums recently proved by Harshitha, Vasuki and Yathirajsharma, using Ramanujan's theory of theta functions, were either already in the literature or can be proved easily by adapting results that can be found in the literature. Also we prove two conjectures given in that paper. After mentioning many other works dealing with identities for various trigonometric sums, we end this paper by describing an automated approach for proving such trigonometric identities.
Inspired by a recent beautiful construction of Armin Straub and Wadim Zudilin, that 'tweaked' the sum of the $s^{th}$ powers of the $n$-th row of Pascal's triangle, getting instead of sequences of numbers, sequences of rational functions, we do the same for general binomial coefficients sums, getting a practically unlimited supply of Apéry limits. While getting what we call "major Apéry miracles", proving irrationality of the associated constants (i.e. the so-called Apéry limits) is very rare, we do get, every time, at least a "minor Apéry miracle" where an explicit constant, defined as an (extremely slowly-converging) limit of some explicit sequence, is expressed as an Apéry limit of some recurrence, with some initial conditions, thus enabling a very fast computation of that constant, with exponentially decaying error.
How many ways, exactly, can a Chess King, always moving forward (i.e. with steps [1,0],[0,1],[1,1]) walk to [100000,200000]? Thanks to the amazing Apagodu-Zeilberger extension of the Almkvist-Zeilberger algorithm, adapted in this article for combinatorial applications, this 104492-digit number, can be computed in less than 33 seconds. But not just this particular number. Many other numbers that come up in enumerative combinatorics, can be computed just as efficiently
Arthur Cayley famously proved that there are n to the power n-2 labeled trees on n vertices. Here we go much further and show how to enumerate, fully automatically, labeled trees such that every vertex has a number of neighbors that belongs to a specified finite set, and also count trees where the number of neighbors is not allowed to be in a given finite set. We also give detailed statistical analysis, and show that in the sample space of labeled trees with n vertices, the random variable "number of vertices with d neighbors" is asymptotically normal, and for any different degrees, are jointly asymptotically normal, but of course, not independently so (except for the pair (1,3), i.e. the number of leaves and the number of degree-3 vertices, where there are asymptotically independent). We also give new proofs to Amram Meir and John Noon's expressions for the limiting expectation and variance for these, and derive an explicit expression for the covariance.
Using recurrences gotten from the Apagodu-Zeilberger Multivariate Almkvist-Zeilberger algorithm we present a linear-time, and constant-space, algorithm to compute the general mixed moments of the k-variate general normal distribution, with any covariance matrix, for any specific k. Besides their obvious importance in statistics, these numbers are also very significant in enumerative combinatorics, since they count in how many ways, in a species with k different genders, a bunch of individuals can all get married, keeping track of the different kinds of heterosexual marriages. We completely implement our algorithm (with an accompanying Maple package, MVNM.txt) for the bivariate and trivariate cases (and hence taking care of our own 2-sex society and a putative 3-sex society), but alas, the actual recurrences for larger k took too long for us to compute. We leave them as computational challenges.