The Importance of Avoiding Satisfaction of Search Bias in a Case of Concurrent Calcaneonavicular Coalition and Osteoid Osteoma

We report a rare case of concurrent calcaneonavicular osseous coalition and osteoid osteoma of the navicular. While each pathology is relatively common, the combination of the two in one foot is rare and as such can present a unique challenge to imaging interpretation. This case reinforces the importance of the concept of “satisfaction of search.”

Best-First Beam Search

Decoding for many NLP tasks requires an effective heuristic algorithm for approximating exact search because the problem of searching the full output space is often intractable, or impractical in many settings. The default algorithm for this job is beam search—a pruned version of breadth-first search. Quite surprisingly, beam search often returns better results than exact inference due to beneficial search bias for NLP tasks. In this work, we show that the standard implementation of beam search can be made up to 10x faster in practice. Our method assumes that the scoring function is monotonic in the sequence length, which allows us to safely prune hypotheses that cannot be in the final set of hypotheses early on. We devise effective monotonic approximations to popular nonmonontic scoring functions, including length normalization and mutual information decoding. Lastly, we propose a memory-reduced variant of best-first beam search, which has a similar beneficial search bias in terms of downstream performance, but runs in a fraction of the time.

Manipulações estatísticas e anomalias visuais: design de visualização de dados e reconhecimento de vieses estatísticos | Statistical manipulations and visual anomalies: data visualization design and statistical bias recognition

O mundo contemporâneo é caracterizado por um amplo volume de informações produzidas. Contudo, proceder a seleção e leitura dessas informações por meio de relatos de pesquisa ou de notícias ainda é um desafio. Entre os obstáculos presentes se destacam os vieses da informação, originados por tratamentos de jornalistas ou pesquisadores, ou mesmo provocados intencionalmente para subverter a representação da realidade a partir dos dados obtidos. Assim, o presente estudo visa discutir a interpretação de informações visuais em representações gráficas de cálculos estatísticos de modo a contextualizar alguns dos principais recursos visuais de enviesamento de pesquisa. Para tanto, aborda os principais modos de enviesamento em pesquisas a partir das representações da estatística e da visualização de dados e identifica alguns passos nos quais o enviesamento se traduz em informações visuais. A partir do levantamento realizado, sugere-se que a compreensão visual dos recursos de visualização de dados pode ao menos instigar a indagação do leitor acerca do possível viés.*****The contemporary world is characterized by a large volume of produced information. However, selecting and reading this information through research reports or news is still a challenge. Among the present obstacles stand out the information bias, originated by treatments of journalists or researchers, or even intentionally provoked to subvert the representation of reality from the obtained data. Thus, the present study aims to discuss the interpretation of visual information in graphical representations of statistical calculations in order to contextualize some of the main visual bias features of research. To this end, it addresses the main modes of search bias from statistical representations and data visualization and identifies some steps in which bias translates into visual information. From the study, it is suggested that the visual understanding of data visualization resources may at least instigate the reader's question about the possible bias.

The Role of Innovation in the Analysis of Abuse of Dominance in Digital Markets: The Analysis of Chosen Practices of Google Search

Competition investigations in digital markets focus increasingly on future markets, and incentives to invest and innovate play here a larger role than in traditional "brick and mortar" industries. the article analyses the role of innovation in cases of abuse of dominance in digital markets on two levels. The first level involves the strength of incentives to invest and innovate of a digital monopolist - would he have less or more incentives to innovate and prefer to resort to practices that foreclose his competitors or leverage his market power to adjacent markets to keep its dominant position on the market? The second level identifies concrete phases of the competition analysis in which innovation considerations are contemplated in digital markets, such as objective justifications or assessing the effect of the practice on consumer welfare. Toe analysis of the role of innovation in the assessment of alleged anticompetitive abuses is conducted on the example of two concerns expressed by the EC in recent investigations into practices of Google Search, namely (1) search bias and (2) restrictions on portability of advertising data to competing advertising platforms.

Enhancing Island Model Genetic Programming by Controlling Frequent Trees

In evolutionary computation approaches such as genetic programming (GP), preventing premature convergence to local minima is known to improve performance. As with other evolutionary computation methods, it can be difficult to construct an effective search bias in GP that avoids local minima. In particular, it is difficult to determine which features are the most suitable for the search bias, because GP solutions are expressed in terms of trees and have multiple features. A common approach intended to local minima is known as the Island Model. This model generates multiple populations to encourage a global search and enhance genetic diversity. To improve the Island Model in the framework of GP, we propose a novel technique using a migration strategy based on textit f requent trees and a local search, where the frequent trees refer to subtrees that appear multiple times among the individuals in the island. The proposed method evaluates each island by measuring its activation level in terms of the fitness value and how many types of frequent trees have been created. Several individuals are then migrated from an island with a high activation level to an island with a low activation level, and vice versa. The proposed method also combines strong partial solutions given by a local search. Using six kinds of benchmark problems widely adopted in the literature, we demonstrate that the incorporation of frequent tree information into a migration strategy and local search effectively improves performance. The proposed method is shown to significantly outperform both a typical Island Model GP and the aged layered population structure method.