Entropy conditions for subsequences of random variables with applications to empirical processes
The authors introduce entropy concepts for classes of subsequences of \(\mathbb{N}\). Given a class \(\mathcal{A}\) of subsequences of \(\mathbb{N}\), for each \(N \geq 1\) let \(\mathcal{A}_N=\{A\cap [1,2,\cdots ,N]:A\in \mathcal{A}\}\), and for \(r \geq 1\) let \(\mathcal{A}_N(r)\) denote the class of sets \(A\in \mathcal{A}_N\) for which \(N2^{-r}<\text{card}\;A\leq N2^{-(r-1)}\). The entropy function of the class \(\mathcal{A}\) is defined by \(\psi(\mathcal{A};N,r):=\text{card}\;\mathcal{A}_N(r)\). Given a sequence \((x_n)\), a generalization of the well-distribution measure of Mauduit and Sárközy is defined by \[ W_N(\mathcal{A})=W_N^{(\mathcal{A})}(x_1,\cdots ,x_N):= \sup_{(p_k)\in \mathcal{A}}\sup_{0\leq t\leq 1}\left| \sum_{p_k\leq N} (1(x_{p_k}\leq t)-t)\right| , \] where \(1(A)\) denotes the indicator function of the set \(A\). The authors show that under the assumption that the entropy function of \(\mathcal{A}\) is not too large, some strong limit theorems hold uniformly over all sequences in \(\mathcal{A}\). For example, the following is proved: Let \((\eta_k)\) be a sequence of independent random variables with uniform distribution over \([0,1)\), let \(\mathcal{A}\) be a class of subsequences of \(\mathbb{N}\) with entropy function \(\psi\) satisfying \[ \psi(\mathcal{A};N,r)\leq \exp (B\cdot 2^{r/2}\log \log N),\: r\geq 0,\: N\geq 10. \] Then with probability 1, \[ 1/4\leq \limsup_{N\to \infty}(N\log \log N)^{-1/2}W_N(\mathcal{A})\leq C \] for some constant \(C\), depending only on the constant \(B\). The authors prove a similar result for a sequence \((n_k\omega)\) mod 1 where the sequence \((n_k)\) satisfies a Hadamard gap conditon. Furthermore, they formulate a similar result for a strictly stationary sequence of random variables satisfying the strong mixing condition.
- scientific article; zbMATH DE number 3923770
- scientific article; zbMATH DE number 4020111
- scientific article; zbMATH DE number 700039
- Absolutely regular empirical processes and universal entropy
- Uniform laws of large numbers for triangular arrays of function-indexed processes under random entropy conditions
- A functional law of the iterated logarithm for empirical distribution functions of weakly dependent random variables
- Central limit theorems for empirical measures
- Invariance principles for sums of Banach space valued random elements and empirical processes
- Measures of pseudorandomness for finite sequences: typical values
- Metric theorems for distribution measures of pseudorandom sequences
- On finite pseudorandom binary sequences VII: The measures of pseudorandomness
- On finite pseudorandom binary sequences. V: On \((n\alpha)\) and \((n^2\alpha)\) sequences
- On finite pseudorandom binary sequences. VI: On \((n^k\alpha)\) sequences
- On pseudorandom \([0,1)\) and binary sequences
- On the pointwise convergence of Fourier series
- Pseudorandom numbers and entropy conditions
- The Law of the Iterated Logarithm for Empirical Distribution
This page was built for publication: Entropy conditions for subsequences of random variables with applications to empirical processes
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2482223)
