Facultatea de Matematică şi Informatică / Faculty of Methematics and Informatics
Permanent URI for this communityhttps://msuir.usm.md/handle/123456789/12
Browse
47 results
Search Results
Item MACHINE GUESSING, PIGEON SUPERSTITION AND TRUST IN AI(CEP USM, 2020) Alvarado, RamónIn this paper I argue that the predictive prowess of artificial intelligence methods, such as machine learning, is not sufficient as an epistemic warrant for us to allocate trust in them. This is because of two reasons: 1) they are opaque in ways that other methodology is not and 2) they fail in ways that other methodology does not. In order to motivate my analysis, I compare the prowess of some machine learning methodology in image recognition to the ability for pattern identification by some animals, such as pigeons, and suggest that while useful in certain practical circumstances, their success does not constitute the kinds of epistemic warrants that we should aim for in more epistemically demanding contexts such as science, medicine, and/or policy-making.Item MEDITATIONS ABOUT THE MYSTIC OF MATHEMATICAL LANGUAGE(CEP USM, 2020) Pagliacci, Mario G.R.Mathematics is the science of measurement of greatness and form. It is fundamental for trying to understand, describe and explain phenomena, especially when they are chaotic and complex. Mathematical language is not true or false for itself. It is neutral in respect of truth or falsity of the described phenomena. However, it can give a distorted representation of reality when it is utilized in not correct way, by doing conceptual forcing or large approximations. Quantitative language can suggest in inexpert minds that anything is expressed in quantitative statement is more credible than qualitative one. The mystic conception that mathematical language is synonym of truth is confuted.Item SISTEME SUPORT PENTRU DECIZII APLICATE ÎN GESTIONAREA SIGURANŢEI RUTIERE(CEP USM, 2021) Bulmaga, ViorelOdată cu utilizarea sistemelor suport pentru decizii (SSD) întru susţinerea deciziilor manageriale în cadrul organizării proceselor de producere, un rol important revine aplicării SSD în abordarea problemei privind managementul siguranţei publice sau al situaţiilor de urgenţă. A fost studiată aplicabilitatea SSD în domeniul managementului siguranţei rutiere pentru a fundamenta deciziile manageriale, în vederea optimizării raportului cost-eficienţă pe următoarele segmente: (1) Identificarea şi clasificarea sectoarelor periculoase pe arterele rutiere, determinând modul şi termenele de intervenţie în proiectarea infrastructurii rutiere pentru a reduce pericolul acestor sectoare; (2) Sprijinirea în timp real a deciziilor conducătorului de vehicul, în baza informaţiilor interne (parametrii vehiculului) şi externe (mesaje privind infrastructura rutieră), pentru a oferi avertismente sau recomandări utile cu privire la pericolele potenţiale ale mediului de conducere; (3) Evaluarea amplorii şi clasificarea situaţiilor de urgenţă în cazul accidentelor de circulaţie, pentru implicarea forţei de muncă şi a resurselor serviciilor de salvare. Urmând exemplele studiate, a fost dedusă posibilitatea aplicării unei reţele neuronale artificiale în scopul de comparare, uniformizare şi de export de date din/în tabele de date cu seturi de variabile şi valori de format diferit (definite diferit). Acest instrument permite realizarea practică a unui Sistem suport pentru decizii aplicat întru soluţionarea unei astfel de probleme nestructurate, precum exportul datelor din Sistemul informaţional automatizat „Registrul de stat al accidentelor rutiere” în formatul CADaS al Uniunii Europene.Item PARALLEL ALGORITHM TO SOLVING 2D BLOCK-CYCLIC PARTITIONED BIMATRIX GAMES(CEP USM, 2021) Hâncu, Boris; Cataranciuc, EmilThe article presents a theoretical and practical study of the ways of determining solutions in bimatrix games divided into blocks of submatrices using 2D block-cyclic matrix dividing and distribution algorithm. The proved theorems represent the foundation on which the bimatrix game solution can be built using the sub-games solutions generated by the 2D-cyclic matrix distribution algorithm.Item TEHNOLOGII INFORMAŢIONALE DE COMUNICARE: ÎNDRUMAR PENTRU LUCRĂRI DE LABORATOR(CEP USM, 2020) Calmîş, Elena; Braguţa, Galina; Buzatu, Radu; Leonte, Stanislav; Ţurcanu, CălinSetul de lucrări de laborator acoperă integral competenţele şi obiectivele generale prevăzute în curriculumul disciplinei TIC, elaborat în cadrul Departamentului Matematică, Facultatea de Matematică şi Informatică (USM) şi destinată studenţilor anului I, Ciclul I Licenţă, ai facultăţilor de Drept, Istorie şi Filosofie, Psihologie şi Ştiinţe ale Educaţiei, Litere, precum şi tuturor celor care doresc să se iniţieze în utilizarea TIC.Item LOCALIZATION OF SINGULAR POINTS OF MEROMORPHIC FUNCTIONS BASED ON INTERPOLATION BY RATIONAL FUNCTIONS(2021) Capcelea, Maria; Capcelea, TituIn this paper we examine two algorithms for localization of singular points of meromorphic functions. Both algorithms apply approximation by interpolation with rational functions. The first one is based on global interpolation and gives the possibility to determine the singular points of the function on a domain that includes a simple closed contour on which the values of the function are known. The second algorithm, based on piecewise interpolation, establishes the poles and the discontinuity points on the contour where the function values are given.Item MAXIMUM NONTRIVIAL CONVEX COVER NUMBER OF JOIN AND CORONA OF GRAPHS(2021) Buzatu, RaduLet G be a connected graph. We say that a set S ⊆ X(G) is convex in G if, for any two vertices x, y ∈ S, all vertices of every shortest path between x and y are in S. If 3 ≤ |S| ≤ |X(G)| − 1, then S is a nontrivial set. The greatest p ≥ 2 for which there is a cover of G by p nontrivial and convex sets is the maximum nontrivial convex cover number of G. In this paper, we determine the maximum nontrivial convex cover number of join and corona of graphs.Item NUMERICAL SIMULATION OF NONLINEAR PROCESSES IN SEMICONDUCTOR DEVICES WITH THE APPLICATION OF THE NEWTON’S METHOD FOR LINEARIZATION(2020) Sprincean, GalinaThis article relates to the use of Newton’s method and Scharfetter– Gummel scheme, to linearize and discretize the equations, for numerical modeling of nonlinear processes in semiconductor devices. The mathematical model of the problem represents a system of nonlinear differential equations, in the unknowns ϕ–electrostatic potential, n, p–the concentrations of electrons and holes, respectively. The problem is further complicated by the fact that the boundary conditions are of two types: the Dirichlet conditions and the Neumman conditions, which act on different portions of the boundary. The subproblems that were solved in this paper: linearization of nonlinear differential equations, using Newton’s method; discretization of equations, using Scharfetter–Gummel scheme. The obtained systems have five diagonal and nonsymmetrical matrices. The numerical method of Bi–Conjugate Gradients was used to solve the systems.Item ALMOST PERIODIC AND ALMOST AUTOMORPHIC SOLUTIONS OF MONOTONE DIFFERENTIAL EQUATIONS WITH A STRICT MONOTONE FIRST INTEGRAL(2020) Ceban, DavidThe paper is dedicated to the study of problem of Poisson stability (in particular periodicity, quasi-periodicity, Bohr almost periodicity, almost automorphy, Levitan almost periodicity, pseudo-periodicity, almost recurrence in the sense of Bebutov, recurrence in the sense of Birkhoff, pseudo-recurrence, Poisson stability) and asymptotical Poisson stability of motions of monotone non-autonomous differential equations which admit a strict monotone first integral. This problem is solved in the framework of general non-autonomous dynamical systems.Item CALCULUL PROPOZIȚIILOR ȘI LOGICA PREDICATELOR: NOTE DE CURS(CEP USM, 2021) Buzatu, Radu; Novac, Ludmila; Cucu, IonLucrarea este direcționată spre formarea competențelor specifice disciplinei Logica Matematică și Teoria Mulțimilor la studenții de la specialitățile de Informatică, Informatică Aplicată și Tehnologii Informaționale de la facultățile: Matematică și Informatică; Fizica și Inginerie.