Yahoo!

About: Yahoo! is a company organization based out in London, United Kingdom. It is known for research contribution in the topics: Population & Web search query. The organization has 26749 authors who have published 29915 publications receiving 732583 citations. The organization is also known as: Yahoo! Inc. & Maudwen-Yahoo! Inc.

Scalable triangulation-based logo recognition

Abstract: We propose a scalable logo recognition approach that extends the common bag-of-words model and incorporates local geometry in the indexing process. Given a query image and a large logo database, the goal is to recognize the logo contained in the query, if any. We locally group features in triples using multi-scale Delaunay triangulation and represent triangles by signatures capturing both visual appearance and local geometry. Each class is represented by the union of such signatures over all instances in the class. We see large scale recognition as a sub-linear search problem where signatures of the query image are looked up in an inverted index structure of the class models. We evaluate our approach on a large-scale logo recognition dataset with more than four thousand classes.

Risk factors in redisplacement of distal radial fractures in children.

Abstract: Background: The causes of redisplacement following closed treatment of distal metaphyseal radial fractures in children are still controversial. Various risk factors and radiographic indices have been suggested to predict redisplacement. The aims of this study were to prospectively identify the causes of redisplacement and to test the accuracy of previously described radiographic indices and our new method, the “three-point index.” Methods: This prospective study included seventy-five displaced or severely angulated distal radial fractures in seventy-four children under the age of fifteen years. Age, gender, initial complete displacement of the radius, an associated ulnar fracture, the accuracy of the reduction, the maximum degree of obliquity of the fracture line in the sagittal or coronal plane, and the distance to the physis were examined as possible risk factors. Logistic regression analysis was utilized to search for risk factors. We also calculated the cast index, padding index, Canterbury index, gap index, and three-point index on the radiographs of each reduction. The sensitivity, specificity, negative predictive value, and positive predictive value were calculated for each test. Results: Initial complete displacement and the degree of obliquity of the fracture were the most important risk factors for redisplacement. Fractures that were completely displaced initially were 11.7 times more likely to redisplace than were angulated but incompletely displaced fractures. A 20° oblique fracture was 4.9 times more likely to redisplace and a 30° oblique fracture was 10.9 times more likely to redisplace than was a 0° true transverse fracture. The three-point index was superior to the other radiographic indices for predicting redisplacement, with a sensitivity of 94.7%, a specificity of 95.2%, a negative predictive value of 98.4%, and a positive predictive value of 85.7%. The gap index was the next-best measure, but it had a sensitivity of 63.2%, a specificity of 76.2%, a negative predictive value of 87.3%, and a positive predictive value of 44.4%. Conclusions: Initial complete displacement and the degree of obliquity of the fracture line are the dominant factors affecting redisplacement. Our new radiographic index, the three-point index, should be used to predict redisplacement and assess the quality of the cast treatment of these fractures. Level of Evidence: Prognostic Level I. See Instructions to Authors for a complete description of levels of evidence.

Intacs in advanced keratoconus.

Abstract: Purpose: To evaluate the surgical outcomes of microthin intracorneal ring segment (INTACS) implantation in advanced keratoconus. Methods: INTACS implantation was performed in eyes with advanced keratoconus. The main outcome measures were uncorrected visual acuity, best spectacle-corrected visual acuity, change in mean refractive spherical equivalent (MRSE), and keratometry. Results: Intacswere implanted in14 eyes withadvanced keratoconus. At 6 months, uncorrected visual acuity improved from 0.05 6 0.08 to 0.16 6 0.11 (P , 0.05), and best spectacle-corrected visual acuity improved from 0.50 6 0.23 to 0.67 6 0.00 (P = 0.01). The spherical refractive error improved from 26.68 D 6 6.44 to 23.11 D 6 3.08 (P = 0.03), whereas the cylindrical refractive error improved from 24.89 D 6 1.91 to 23.64 D 6 1.27 (P = 0.04). The MRSE reduced from 29.13 D 6 5.62 to 24.93 D 6 3.19 (P = 0.01), and the average keratometry decreased from 53.01 D 6 3.70 to 49.42 D 6 3.79 (P , 0.05). The results were stable from 6 months to 1 year. The procedure showed 100% safety, and more than 60% tolerated contact lenses. Younger age, male sex, and minimum central pachymetry of more than 400 mm seemed to be associated with better outcomes. Conclusions: In this series, Intacs provided good results with respect to visual acuity, corneal topography, and MRSE in eyes with advanced keratoconus without major complications or the need for segment explanation. INTACS is potentially a safe and efficacious treatment option in the management of advanced keratoconus.

{Tight Regret Bounds for Stochastic Combinatorial Semi-Bandits}

Abstract: A stochastic combinatorial semi-bandit is an online learning problem where at each step a learning agent chooses a subset of ground items subject to constraints, and then observes stochastic weights of these items and receives their sum as a payoff. In this paper, we close the problem of computationally and sample efficient learning in stochastic combinatorial semi-bandits. In particular, we analyze a UCB-like algorithm for solving the problem, which is known to be computationally efficient; and prove O(KL(1/)logn) and O( p KLnlogn) upper bounds on its n-step regret, where L is the number of ground items, K is the maximum number of chosen items, and is the gap between the expected returns of the optimal and best suboptimal solutions. The gapdependent bound is tight up to a constant factor and the gap-free bound is tight up to a polylogarithmic factor.

Limited individual attention and online virality of low-quality information

Abstract: Why does low-quality information go viral? A stylized model of social media predicts that under real-world conditions of high information load and limited attention, low- and high-quality information are equally likely to go viral.