arXiv subjects

Mrinal Kanti Roychowdhury

Publications and source records attributed to Mrinal Kanti Roychowdhury.

48 records · Page 3Linked to original sources

arXiv2019

Topological pressure and fractal dimensions of cookie-cutter-like sets

The cookie-cutter-like set is defined as the limit set of a sequence of classical cookie-cutter mappings. For this cookie-cutter set it is shown that the topological pressure function exists, and that the fractal dimensions such as the Hausdorff dimension, the packing dimension and the box-counting dimension are all equal to the unique zero $h$ of the pressure function. Moreover, it is shown that the $h$-dimensional Hausdorff measure and the $h$-dimensional packing measure are finite and positive.

math.DS↗

arXiv2018

Quantization dimension for infinite conformal iterated function systems

The quantization dimension function for an $F$-conformal measure $m_F$ generated by an infinite conformal iterated function system satisfying the strong open set condition and by a summable Hölder family of functions is expressed by a simple formula involving the temperature function of the system. The temperature function is commonly used to perform the multifractal analysis, in our context of the measure $m_F$. The result in this paper extends a similar result of Lindsay and Mauldin established for finite conformal iterated function systems [Nonlinearity 15 (2002)].

math.DS↗

arXiv2018

Optimal quantization for the Cantor distribution generated by infinite similutudes

Let $P$ be a Borel probability measure on $\mathbb R$ generated by an infinite system of similarity mappings $\{S_j : j\in \mathbb N\}$ such that $P=\sum_{j=1}^\infty \frac 1{2^j} P\circ S_j^{-1}$, where for each $j\in \mathbb N$ and $x\in \mathbb R$, $S_j(x)=\frac 1{3^{j}}x+1-\frac 1 {3^{j-1}}$. Then, the support of $P$ is the dyadic Cantor set $C$ generated by the similarity mappings $f_1, f_2 : \mathbb R \to \mathbb R$ such that $f_1(x)=\frac 13 x$ and $f_2(x)=\frac 13 x+\frac 23$ for all $x\in \mathbb R$. In this paper, using the infinite system of similarity mappings $\{S_j : j\in \mathbb N\}$ associated with the probability vector $(\frac 12, \frac 1{2^2}, \cdots)$, for all $n\in \mathbb N$, we determine the optimal sets of $n$-means and the $n$th quantization errors for the infinite self-similar measure $P$. The technique obtained in this paper can be utilized to determine the optimal sets of $n$-means and the $n$th quantization errors for more general infinite self-similar measures.

math.DS↗

arXiv2018

High Precision Numerical Computation of Principal Points For Univariate Distributions

Principal points were first introduced by Flury: for a positive integer $n$, $n$ principal points of a random variable are the $n$ points that minimize the mean squared distance between the random variable and the nearest of the $n$ points. In this paper, we determine the $n$ principal points and the corresponding values of mean squared distance for different values of $n$ for some univariate absolutely continuous distributions.

math.PR↗

arXiv2018

Least upper bound of the exact formula for optimal quantization of some uniform Cantor distributions

The quantization scheme in probability theory deals with finding a best approximation of a given probability distribution by a probability distribution that is supported on finitely many points. Let $P$ be a Borel probability measure on $\mathbb R$ such that $P=\frac 12 P\circ S_1^{-1}+\frac 12 P\circ S_2^{-1},$ where $S_1$ and $S_2$ are two contractive similarity mappings given by $S_1(x)=rx$ and $S_2(x)=rx+1-r$ for $0<r<\frac 12$ and $x\in \mathbb R$. Then, $P$ is supported on the Cantor set generated by $S_1$ and $S_2$. The case $r=\frac 13$ was treated by Graf and Luschgy who gave an exact formula for the unique optimal quantization of the Cantor distribution $P$ (Math. Nachr., 183 (1997), 113-133). In this paper, we compute the precise range of $r$-values to which Graf-Luschgy formula extends.

math.DS↗

arXiv2018

Asymptotics of the quantization errors for condensation measures

Let $P:=\frac 1 3 P\circ S_1^{-1}+\frac 13 P\circ S_2^{-1}+\frac 13ν$, where $S_1(x)=\frac 15 x$, $S_2(x)=\frac 1 5 x+\frac 45$ for all $x\in \mathbb R$, and $ν$ be a Borel probability measure on $\mathbb R$ with compact support. Such a measure $P$ is called a condensation measure, or an an inhomogeneous self-similar measure, associated with the condensation system $(\{S_1, S_2\}, (\frac 13, \frac 13, \frac 13), \gn)$. In this paper, we have explicitly calculated optimal quantizers, quantization dimension, and the lower and upper quantization coefficients for an inhomogeneous self-similar measure.

math.DS↗

arXiv2018

Optimal quantization for piecewise uniform distributions

Quantization for a probability distribution refers to the idea of estimating a given probability by a discrete probability supported by a finite number of points. In this paper, firstly a general approach to this process is outlined using independent random variables and ergodic maps; these give asymptotically the optimal sets of $n$-means and the $n$th quantization errors for all positive integers $n$. Secondly two piecewise uniform distributions are considered on $\mathbb R$: one with infinite number of pieces and one with finite number of pieces. For these two probability measures, we describe the optimal sets of $n$-means and the $n$th quantization errors for all $n\in \mathbb N$. It is seen that for a uniform distribution with infinite number of pieces to determine the optimal sets of $n$-means for $n\geq 2$ one needs to know an optimal set of $(n-1)$-means, but for a uniform distribution with finite number of pieces one can directly determine the optimal sets of $n$-means and the $n$th quantization errors for all $n\in \mathbb N$.

math.PR↗

arXiv2017

Optimal quantizers for some absolutely continuous probability measures

The representation of a given quantity with less information is often referred to as `quantization' and it is an important subject in information theory. In this paper, we have considered absolutely continuous probability measures on unit discs, squares, and the real line. For these probability measures the optimal sets of $n$-means and the $n$th quantization errors are calculated for some positive integers $n$.

math.PR↗

arXiv2017

Quantization and centroidal Voronoi tessellations for probability measures on dyadic Cantor sets

Quantization of a probability distribution is the process of estimating a given probability by a discrete probability that assumes only a finite number of levels in its support. Centroidal Voronoi tessellations (CVT) are Voronoi tessellations of a region such that the generating points of the tessellations are also the centroids of the corresponding Voronoi regions. In this paper, we investigate the optimal quantization and the centroidal Voronoi tessellations with $n$ generators for a Borel probability measure $P$ on $\mathbb R$ supported by a dyadic Cantor set generated by two self-similar mappings with similarity ratios $r$, where $0<r\leq \frac{5-\sqrt{17}}2$.

math.DS↗

arXiv2017

Quantization for uniform distributions on equilateral triangles

We approximate the uniform measure on an equilateral triangle by a measure supported on $n$ points. We find the optimal sets of points ($n$-means) and corresponding approximation (quantization) error for $n\leq4$, give numerical optimization results for $n\leq 21$, and a bound on the quantization error for $n\to\infty$. The equilateral triangle has particularly efficient quantizations due to its connection with the triangular lattice. Our methods can be applied to the uniform distributions on general sets with piecewise smooth boundaries.

cs.IT↗

arXiv2015

Quantization dimension for Gibbs-like measures on cookie-cutter sets

In this paper using Banach limit we have determined a Gibbs-like measure $μ_h$ supported by a cookie-cutter set $E$ which is generated by a single cookie-cutter mapping $f$. For such a measure $μ_h$ and $r\in (0, +\infty)$ we have shown that there exists a unique $κ_r \in (0, +\infty)$ such that $κ_r$ is the quantization dimension function of the probability measure $μ_h$, and established its functional relationship with the temperature function of the thermodynamic formalism. The temperature function is commonly used to perform the multifractal analysis, in our context of the measure $μ_h$. In addition, we have proved that the $κ_r$-dimensional lower quantization coefficient of order $r$ of the probability measure is positive.

math.DS↗

arXiv2013

A class of measures and non-stationary fractals, associated to f-expansions

We construct first a class of Moran fractals in R^d with countably many generators and non-stationary contraction rates; at each step n, the contractions depend on n-truncated sequences, and are related to asymptotic letter frequencies. In some cases the sets of contractions may be infinite at each step. We show that the Hausdorff dimension of such a fractal is equal to the zero h of a pressure function. We prove that the dimensions of these sets depend real analytically on the frequencies. Next, we apply the above construction to obtain non-stationary fractals E(x; f) \subset R^d, associated to f-expansions of real numbers x, and study the dependence of these fractals on x. We consider for instance beta-expansions, the continued fraction expansion and other f-expansions. By employing the Ergodic Theorem for invariant absolutely continuous measures and equilibrium measures, and using some probabilities for which the digits become independent random variables, we study the function x \to dim_H(E(x; f)) on the respective set of quasinormal numbers x \in [0; 1). We investigate also another class of fractals \tilde E_f (x) \subset R^d, for which both the non-stationary contraction vectors and the asymptotic frequencies depend on the f-representation of x. We obtain then some properties of the digits of x, related to \tilde E_f (x) and to equilibrium measures.

math.DS↗