No abstract

Cardinality estimation has long been grounded in statistical tools for density estimation. To capture the rich multivariate distributions of relational tables, we propose the use of a new type of high-capacity statistical model: deep autoregressive models. However, direct application of these models leads to a limited estimator that is prohibitively expensive to evaluate for range or wildcard predicates. To produce a truly usable estimator, we develop a Monte Carlo integration scheme on top of autoregressive models that can efficiently handle range queries with dozens of dimensions or more. Like classical synopses, our estimator summarizes the data without supervision. Unlike previous solutions, we approximate the joint data distribution without any independence assumptions. Evaluated on real-world datasets and compared against real systems and dominant families of techniques, our estimator achieves single-digit multiplicative error at tail, an up to 90x accuracy improvement over the second best method, and is space- and runtime-efficient.

Numerous factors are thought to be advantageous for non-native language learning although they are typically investigated in isolation, and the interaction between them is not understood. Firstly, bilinguals are claimed to acquire a third language easier than monolinguals acquire a second. Secondly, closely related languages may be easier to learn. Thirdly, certain phonetic features could be universally more difficult to acquire. We tested these hypotheses used as explanations by having adults learn vocabularies that differentiated words using foreign phonetic contrasts. In Experiment 1, Mandarin–English bilinguals outlearned English monolinguals, and the Mandarin-like (retroflex) artificial language was better learned than the English-like (fricative voicing). In Experiment 2, bilinguals again outlearned English monolinguals for the Mandarin-like artificial language. However, only Korean–English bilinguals showed an advantage for the more difficult Korean-like (lenition) language. Bilinguals, relative to monolinguals, show a general advantage when learning ‘easy’ contrasts, but phonetic similarity to the native language is useful for learning universally ‘difficult’ contrasts.

A key challenge in leveraging data augmentation for neural network training is choosing an effective augmentation policy from a large search space of candidate operations. Properly chosen augmentation policies can lead to significant generalization improvements; however, state-of-theart approaches such as AutoAugment are computationally infeasible to run for the ordinary user. In this paper, we introduce a new data augmentation algorithm, Population Based Augmentation (PBA), which generates nonstationary augmentation policy schedules instead of a fixed augmentation policy. We show that PBA can match the performance of AutoAugment on CIFAR-10, CIFAR-100, and SVHN, with three orders of magnitude less overall compute. On CIFAR-10 we achieve a mean test error of 1.46%, which is a slight improvement upon the current state-of-the-art. The code for PBA is open source and is available at https://github.com/arcelien/pba.