Interpolation that leads to the narrowest intervals and its application to expert systems and intelligent control

In many real-life situations, we want to reconstruct the dependencyy=f(x ₁,…, x_n) from the known experimental resultsx _i ^(k) , y^(k). In other words, we want tointerpolate the functionf from its known valuesy ^(k)=f(x ₁ ^(k) ,…, x _n ^(k) ) in finitely many points $\bar x^{(k)} = (x_1^{(k)} , \ldots ,x_n^{(k)} )$ , 1≤k≤N There are many functions that go through given points. How to choose one of them? The main goal of findingf is to be able to predicty based onx _i. If we getx _i from measurements, then usually, we only getintervals that containx _i. As a result of applyingf, we get an interval y of possible values ofy. It is reasonable to choosef for which the resulting interval is the narrowest possible. In this paper, we formulate this choice problem in mathematical terms, solve the corresponding problem for several simple cases, and describe the application of these solutions to intelligent control.     

Checking If There Exists a Monotonic Function That Is Consistent with the Measurements: An Efficient Algorithm

In many problems in science and engineering ranging from astrophysics to geosciences to financial analysis, we know that a physical quantity y depends on the physical quantity x, i.e., y = f(x) for some function f(x), and we want to check whether this dependence is monotonic. Specifically, finitely many measurements of x_i and y = f(x) have been made, and we want to check whether the results of these measurements are consistent with the monotonicity of f(x). An efficient parallelizable algorithm is known for solving this problem when the values x_i are known precisely, while the values y_i are known with interval uncertainty. In this paper, we extend this algorithm to a more general (and more realistic) situation when both x_i and y_i are known with interval uncertainty.     

Choosing,Agreeing, and Eliminating in Communication Complexity

We consider several questions inspired by the direct-sum problem in (two-party) communication complexity. In all questions, there are k fixed Boolean functions f ₁,…,f _k and each of Alice and Bob has k inputs, x ₁,…,x _k and y ₁,…,y _k , respectively. In the eliminate problem, Alice and Bob should output a vector σ₁,…,σ _k such that f _i (x _i , y _i ) ≠ σ _i for at least one i (i.e., their goal is to eliminate one of the 2 ^k output vectors); in the choose problem, Alice and Bob should return (i, f _i (x _i , y _i )), for some i (i.e., they choose one instance to solve), and in the agree problem they should return f _i (x _i , y _i ), for some i (i.e., if all the k Boolean values agree then this must be the output). The question, in each of the three cases, is whether one can do better than solving one (say, the first) instance. We study these three problems and prove various positive and negative results. In particular, we prove that the randomized communication complexity of eliminate, of k instances of the same function f, is characterized by the randomized communication complexity of solving one instance of f.     

Fast error estimates for indirect measurements: applications to pavement engineering

In many real-life situations, we have the following problem: we want to know the value of some characteristicy that is difficult to measure directly (e.g., lifetime of a pavement, efficiency of an engine, etc.). To estimatey, we must know the relationship betweeny and some directly measurable physical quantitiesx ₁,...,x _n . From this relationship, we extract an algorithmf that allows us, givenx _i , to computey: y=f(x ₁, ...,x _n ). So, we measurex _i , apply an algorithmf, and get the desired estimate. Existing algorithms for error estimate (interval mathematics, Monte-Carlo methods, numerical differentiation, etc.) require computation time that is several times larger than the time necessary to computey=f(x ₁, ...,x _n ). So, if an algorithmf is already time-consuming, error estimates will take too long. In many cases, this algorithmf consists of two parts: first, we usex _i to determine the parametersz _k of a model that describes the measured object, and second, we use these parameters to estimatey. The most time-consuming part is findingz _k ; this is done by solving a system of non-linear equations; usually least squares method is used. We show that for suchf, one can estimate errors repeating this time-consuming part off only once. So, we can compute bothy and an error estimate fory with practically no increase in total computation time. As an example of this methodology, we give pavement lifetime estimates.     

Bounded dilation maps of hypercubes into Cayley graphs on the symmetric group

LetG andH be graphs with |V(G)≤ |V(H)|. Iff:V(G) →V(H) is a one-to-one map, we letdilation(f) be the maximum of dist _H (f x),f(y)) over all edgesxy inG where dist _H denotes distance inH. The construction of maps fromG toH of small dilation is motivated by the problem of designing small slowdown simulations onH of algorithms that were originally designed for the networkG. LetS(n), thestar network of dimension n, be the graph whose vertices are the elements of the symmetric group of degreen, two verticesx andy being adjacent ifx o (1,i) =y for somei. That is,xy is an edge ifx andy are related by a transposition involving some fixed symbol (which we take to be 1). Also letP(n), thepancake network of dimension n, be the graph whose vertices are the elements of the symmetric group of degreen, two verticesx andy being adjacent if one can be obtained from the other by reversing some prefix. That is,xy is an edge ifx andy are related byx o (1,i(2,i-1) ⋯ ([i/2], [i/2]) =y. The star network (introduced in [AHK]) has nice symmetry properties, and its degree and diameter are sublogarithmic as functions of the number of vertices, making it compare favorably with the hypercube network. These advantages ofS(n) motivate the study of how well it can simulate other parallel computation networks, in particular, the hypercube. The concern of this paper is to construct low dilation maps of hypercube networks into star or pancake networks. Typically in such problems, there is a tradeoff between keeping the dilationsmall and simulating alarge hypercube. Our main result shows that at the cost ofO (1) dilation asn→ ∞, one can embed a hypercube of near optimum dimension into the star or pancake networksS(n) orP(n). More precisely, lettingQ (d) be the hypercube of dimensiond, our main theorem is stated below. For simplicity, we state it only in the special case when the star network dimension is a power of 2. A more general result (applying to star networks of arbitrary dimension) is obtained by a simple interpolation. This author's research was done during the Spring Semester 1991, as a visiting professor in the Department of Mathematics and Statistics at Miami University.     

Quadrature rules for the integration of the product of two oscillatory functions with different frequencies

We construct quadrature rules for the efficient computation of the integral of a product of two oscillatory functions y₁(x) and y₂(x), where , and the functions f_i,j(x) are smooth. The weights are evaluated by the exponential fitting technique of Ixaru [Comput. Phys. Comm. 105 (1997) 1-19], which is now extended to cover the case of two frequencies. We give a numerical illustration on how the new rules compare for accuracy with the one-frequency dependent rules and with the classical ones.     

Cylindrical decomposition for systems transcendental in the first variable

We present a generalization of the Cylindrical Algebraic Decomposition (CAD) algorithm to systems of equations and inequalities in functions of the form p(x,f₁(x),…,f_m(x),y₁,…,y_n), where p∈Q[x,t₁,…,t_m,y₁,…,y_n] and f₁(x),…,f_m(x) are real univariate functions such that there exists a real root isolation algorithm for functions from the algebra Q[x,f₁(x),…,f_m(x)]. In particular, the algorithm applies when f₁(x),…,f_m(x) are real exp-log functions or tame elementary functions.     

Two point hermite approximations for the solution of linear initial value and boundary value problems

Application of an idea originally due to Ch. Hermite allows the derivation of an approximate formula for expressing the integral ∫xixi?1y(x)dx as a linear combination of y(xi?1), y(xi), and their derivatives y(v)(xi?1) up to order v = α and y(v)(xi) up to order v = β. In addition to this integro-differential form a purely differential form of the 2-point Hermite approximation will be derived. Both types will be denoted by Hαβ-approximation. It will be shown that the well-known Obreschkoff-formulas contain no new elements compared to the much older Hαβ-method.The Hαβ-approximation will be applied to the solution of systems of ordinary differential equations of the type y'(x) = M(x)y(x) + q(x), and both initial value and boundary value problems will be treated. Function values at intermediate points x? (xi?1, xi) are obtained by the use of an interpolation formula given in this paper.An advantage of the Hαβ-method is the fact that high orders of approximation (α, β) allow an increase in step size hi. This will be demonstrated by the results of several test calculations.     

A recursive and a grammatical characterization of the exponential-time languages

We characterize the class of all languages which are acceptable in exponential time by means of recursive and grammatical methods. (i) The class of all languages which are acceptable in exponential time is uniquely characterized by the class of all (0-1)-functions which can be generated, starting with the initial functions of the Grzegorczyk-class $\text{E}$², by means of subtitution and limited recursion of the form f(x, y + 1) = h(x, y), f(x, y), f(x, l(x, y))), l(x, y) ? y. (ii) The class of all languages which are acceptable in exponential time is equal to the class of all languages generated by context-sensitive grammars with context-free control sets.     

Approximate Solutions of Polynomial Equations

In this paper, we introduce “approximate solutions" to solve the following problem: given a polynomial F(x, y) over Q, where x represents an n -tuple of variables, can we find all the polynomials G(x) such that F(x, G(x)) is identically equal to a constant c in Q ? We have the following: let F(x, y) be a polynomial over Q and the degree of y in F(x, y) be n. Either there is a unique polynomial g(x)  ∈ Q [ x ], with its constant term equal to 0, such that F(x, y)  = ∑_j = 0ⁿc_j(y − g(x))^jfor some rational numbers c_j, hence, F(x, g(x)  + a)  ∈ Q for all a ∈ Q, or there are at most t distinct polynomials g₁(x),⋯ , g_t(x), t ≤ n, such that F(x, g_i(x))  ∈ Q for 1  ≤ i ≤ t. Suppose that F(x, y) is a polynomial of two variables. The polynomial g(x) for the first case, or g₁(x),⋯ , g_t(x) for the second case, are approximate solutions of F(x, y), respectively. There is also a polynomial time algorithm to find all of these approximate solutions. We then use Kronecker’s substitution to solve the case of F(x, y).     

Numerische Lösung von Anfangswertaufgaben gewöhnlicher Differentialgleichungen n-ter Ordnung

In the present paper a new method is given for the numerical treatment of the initial problemsy ⁽ⁿ⁾=f(x,y,y′, ...,y ^(n?1),y ⁽ⁱ⁾ (x _o )=y _o ⁽ⁱ⁾ , i=0, 1, ...,n?1. This method is an one-step process of order four. For a class of linear differential equations the exact solution is obtained. Moreover some numerical results are presented.     

Asymptotic behaviour of the solutions of second order functional differential equations

This paper presents integral criteria to determine the asymptotic behaviour of the solutions of second order nonlinear differential equations of the type y^″(x)+q(x)f(y(x))=0, with q(x)>0 and f(y) odd and positive for y>0, as x tends to +∞. It also compares them with the results obtained by Chanturia (1975) in [11] for the same problem.     

Existence of solutions to singular integral equations

Nonnegative solutions are established for singular integral equations of the form y(t) = h(t) + ∫^T₀ k(t, s)f(s, y(s)) ds for t ∈ [0, T]. Here f may be singular at y = 0.     

On derivative-based global sensitivity criteria

A mathematical model f(x) given in the unit n-dimensional cube, where x = (x ₁, ..., x _n ), is considered. How can one estimate the global sensitivity of f(x) with respect to x _i ? If f(x) ∈ L ₂, global sensitivity indices help answer this question. Being less reliable, derivative-based sensitivity criteria are sometimes easier-to-compute. In this work, a new derivative-based global sensitivity criterion is compared to the respective global sensitivity index. These estimates are proven to coincide in the particular case when f(x) linearly depends on x _i . However, Monte Carlo approximations to the derivative-based criterion converge faster. Thus, the global derivative-based sensitivity criterion can prove useful when f(x) depends on x _i almost linearly. It can be also used to find nonessential variables x _i .     

Diagonal polynomials and diagonal orders on multidimensional lattices

HereR andN denote the real numbers and the nonnegative integers, respectively. Alsos(x)=x ₁+···+x _n whenx=(x ₁, …,x _n) inR ⁿ. A mapf:R ⁿ →R is call adiagonal function of dimensionn iff|N ⁿ is a bijection ontoN and, for allx, y inN ⁿ, f(x)<f(y) whens(x)<s(y). Morales and Lew [6] constructed 2 ⁿ⁻² inequivalent diagonal polynomial functions of dimensionn for eachn>1. Here we use new combinatorial ideas to show that numberd _n of such functions is much greater than 2 ⁿ⁻² forn>3. These combinatorial ideas also give an inductive procedure to constructd _n+1 diagonal orderings of {1, …,n}.     

Curve generation of implicit functions by incremental computers

The conventional numerical solution of an implicit function f(x, y) = 0 is substantially complicated for calculating by any computer. We propose a new method representing the argument of the implicit function as a unary function of a parameter, t, if the continuous and unique solution of f(x, y) = 0 exists. The total differential $\text{d}\text{f}\text{d}\text{t}$ constitutes simultaneous differential equations of which the solution about x and y is unique. The Newton-Raphson method must be used to calculate the values near singular points of an implicit function and then the sign of dt has to be decided according to four special cases. Incremental computers are suitable for curve generation of implicit functions by the new method, because the incremental computer can perform more complex algorithms than the analog computer and can calculate faster than the digital computer. This method is easily applicable to curve generation in three-dimensional space.     

Some procedures for function approximation based on the use of sample data and their application in heuristic methods for solving practical problems

Let f(x) be a member of a set of functions over a probability space. Samples of f(x) are 2-tuples (xⁱ,f(xⁱ) where xⁱ is a sample of the random variable X and f(xⁱ) is a sample of f(x) at x = xⁱ. Some procedures and analysis are presented for the approximation of such functions by systems of orthonormal functions. The approximations are based on the data samples. The analysis includes the case of error in the measurement of f(xⁱ). The properties of the expected square error in the approximation are examined for a number of different estimators for the coefficients in the expansion and these well-behaved and easily analyzed estimators are compared to those obtained using the method of least squares. The effectiveness of different sets of basis functions, those involved in the Karhunen-Loeve expansion and others, can be compared and an approach is suggested to adaptive basis selection in order to select that basis which is most efficient in approximating the particular function under examination. The connection between results and applications are discussed in the introduction and conclusion.     

On multiplication in algebraic extension fields

In this paper we characterize all algorithms for obtaining the coefficients of (Σ^n?1_i=0x_iuⁱ)(Σ^n?1_i=0y_iuⁱ) mod P(u), where P(u) is an irreducible po lynomial of degree n, which use 2n ? 1 multiplications. It is shown that up to equivalence, all such algorithms are obtainable by first obtaining the coefficients of the product of two polynomials, and then reducing modulo the irreducible polynomial.     

Bounded solutions of nonlinear difference equations with advanced arguments

Sufficient conditions of existence and uniqueness of α-bounded and bounded solutions to the difference equation with advancedd arguments x(n + 1) = A(n)x(n) + B(n)x(σ₁(n)) + f(n, x(n), x(σ₂(n)), σ_i(n) ⩾ n + 1, i = 1,2, are given. It is proven that under certain conditions it is possible to find positive numbers R, μ, such that from every initial condition ξ satisfying ∥ξ∥ ⩽ R, a unique bounded solution, belonging to the ball ∥x∥ ⩽ μ, starts.     

Computing monodromy via continuation methods on random Riemann surfaces

We consider a Riemann surface X defined by a polynomial f(x,y) of degree d, whose coefficients are chosen randomly. Hence, we can suppose that X is smooth, that the discriminant δ(x) of f has d(d−1) simple roots, Δ, and that δ(0)≠0, i.e. the corresponding fiber has d distinct points {y₁,…,y_d}. When we lift a loop 0∈γ⊂C−Δ by a continuation method, we get d paths in X connecting {y₁,…,y_d}, hence defining a permutation of that set. This is called monodromy.Here we present experimentations in Maple to get statistics on the distribution of transpositions corresponding to loops around each point of Δ. Multiplying families of “neighbor” transpositions, we construct permutations and the subgroups of the symmetric group they generate. This allows us to establish and study experimentally two conjectures on the distribution of these transpositions and on transitivity of the generated subgroups.Assuming that these two conjectures are true, we develop tools allowing fast probabilistic algorithms for absolute multivariate polynomial factorization, under the hypothesis that the factors behave like random polynomials whose coefficients follow uniform distributions.