Amazon.com

About: Amazon.com is a company organization based out in Seattle, Washington, United States. It is known for research contribution in the topics: Computer science & Service (business). The organization has 13363 authors who have published 17317 publications receiving 266589 citations.

Eventually consistent

Abstract: Building reliable distributed systems at a worldwide scale demands trade-offs between consistency and availability.

Collaborative recommendations using item-to-item similarity mappings

Abstract: A recommendations service recommends items to individual users based on a set of items that are known to be of interest to the user, such as a set of items previously purchased by the user. The service is used to recommend products to users of a merchant's Web site (30). The service generates the recommendations using a previously-generated table (60) which maps items (62) to lists (64) of 'similar' items. The similarities reflected by the table (60) are based on the collective interests of the community of users. To generate personal recommendations, the service retrieves from the table (60) the similar items lists (64) corresponding to the items known to be of interest to the user. These similar items lists (64) are appropriately combined into a single list, which is then sorted and filtered to generate a list of recommended items. Also disclosed are various methods for using the current and/or past contents of a user's electronic shopping cart to generate recommendations.

VL2: a scalable and flexible data center network

Abstract: To be agile and cost effective, data centers must allow dynamic resource allocation across large server pools. In particular, the data center network should provide a simple flat abstraction: it should be able to take any set of servers anywhere in the data center and give them the illusion that they are plugged into a physically separate, noninterfering Ethernet switch with as many ports as the service needs. To meet this goal, we present VL2, a practical network architecture that scales to support huge data centers with uniform high capacity between servers, performance isolation between services, and Ethernet layer-2 semantics. VL2 uses (1) flat addressing to allow service instances to be placed anywhere in the network, (2) Valiant Load Balancing to spread traffic uniformly across network paths, and (3) end system--based address resolution to scale to large server pools without introducing complexity to the network control plane. VL2's design is driven by detailed measurements of traffic and fault data from a large operational cloud service provider. VL2's implementation leverages proven network technologies, already available at low cost in high-speed hardware implementations, to build a scalable and reliable network architecture. As a result, VL2 networks can be deployed today, and we have built a working prototype. We evaluate the merits of the VL2 design using measurement, analysis, and experiments. Our VL2 prototype shuffles 2.7 TB of data among 75 servers in 395 s---sustaining a rate that is 94% of the maximum possible.

Bag of Tricks for Image Classification with Convolutional Neural Networks

Abstract: Much of the recent progress made in image classification research can be credited to training procedure refinements, such as changes in data augmentations and optimization methods. In the literature, however, most refinements are either briefly mentioned as implementation details or only visible in source code. In this paper, we will examine a collection of such refinements and empirically evaluate their impact on the final model accuracy through ablation study. We will show that, by combining these refinements together, we are able to improve various CNN models significantly. For example, we raise ResNet-50's top-1 validation accuracy from 75.3% to 79.29% on ImageNet. We will also demonstrate that improvement on image classification accuracy leads to better transfer learning performance in other application domains such as object detection and semantic segmentation.

Hyperdominance in the Amazonian Tree Flora

Abstract: The vast extent of the Amazon Basin has historically restricted the study of its tree communities to the local and regional scales. Here, we provide empirical data on the commonness, rarity, and richness of lowland tree species across the entire Amazon Basin and Guiana Shield (Amazonia), collected in 1170 tree plots in all major forest types. Extrapolations suggest that Amazonia harbors roughly 16,000 tree species, of which just 227 (1.4%) account for half of all trees. Most of these are habitat specialists and only dominant in one or two regions of the basin. We discuss some implications of the finding that a small group of species—less diverse than the North American tree flora—accounts for half of the world’s most diverse tree community.