Prioritizing predictability over efficiency, adapting data partitioning to traffic, and continuous verification are a few of the principles that help ensure stability, availability, and efficiency. | Continue reading
CAV keynote lecture by the director of applied science for AWS Identity explains how AWS is making the power of automated reasoning available to all customers. | Continue reading
Contiguous parameter management and prefetched activation offloading expand the MiCS tool kit. | Continue reading
To mark the occasion of the eighth Federated Logic Conference (FloC), Amazon’s Byron Cook, Daniel Kröning, and Marijn Heule discussed automated reasoning’s prospects. | Continue reading
Paper presents a criterion for halting the hyperparameter optimization process. | Continue reading
Syne Tune supports multiple backends, single-fidelity and multi-fidelity (early-exit) optimization algorithms, and hyperparameter transfer learning. | Continue reading
Amazon DynamoDB is a NoSQL cloud database service that provides consistent performance at any scale. Hundreds of thousands of customers rely on DynamoDB for its fundamental properties: consistent performance, availability, durability, and a fully managed serverless experience. In … | Continue reading
Amazon ICML paper proposes information-theoretic measurement of quantitative causal contribution. | Continue reading
Fun visual essays explain key concepts of machine learning. | Continue reading
Amazon’s TabTransformer model is now available through SageMaker JumpStart and the official release of the Keras open-source library. | Continue reading
With the increasing adoption of machine learning (ML) models and systems in high-stakes settings across different industries, guaranteeing a model’s performance after deployment has become crucial. Monitoring models in production is a critical aspect of ensuring their continued p … | Continue reading
Automated-reasoning method enables the calculation of tight bounds on the use of resources — such as computation or memory — that results from code changes. | Continue reading
Speed of delivery is critical for the success of e-commerce platforms. Faster delivery promise to the customer results in increased conversion and revenue. There are typically two mechanisms to control the delivery speed - a) replication of products across warehouses, and b) air- … | Continue reading
New method would enable BERT-based natural-language-processing models to handle longer text strings, run in resource-constrained settings — or sometimes both. | Continue reading
A/B tests, also known as online controlled experiments, have been used at scale by data-driven enterprises to guide decisions and test innovative ideas. Meanwhile, non-stationarity, such as the time-of-day effect, can commonly arise in various business metrics. We show that inade … | Continue reading
New features go beyond conventional effect estimation by attributing events to individual components of complex systems. | Continue reading
Reformulating the mapping problem to take advantage of sequence-to-sequence Transformers improves performance by an average of 15%. | Continue reading
Amazon fulfillment centers use thousands of mobile robots. To keep products moving, Amazon Robotics researchers have crafted unique solutions. | Continue reading
Training mixed-domain translation models is a complex task that demands tailored architectures and costly data preparation techniques. In this work, we leverage federated learning (FL) in order to tackle the problem. Our investigation demonstrates that with slight modifications i … | Continue reading
Amazon Web Services (AWS) is a comprehensive and broadly adopted cloud provider. AWS SDKs provide access to AWS services through API endpoints. However, incorrect use of these APIs can lead to code defects, crashes, performance issues, and other problems. AWS best practices are a … | Continue reading
Deep learning to produce invariant representations, estimations of sensor reliability, and efficient map representations all contribute to Astro’s superior spatial intelligence. | Continue reading
Amazon researchers optimize the distributed-training tool to run efficiently on the Elastic Fabric Adapter network interface. | Continue reading
New phase estimation technique reduces qubit count, while learning framework enables characterization of noisy quantum systems. | Continue reading
Amazon joins the US DOE’s Bio-Optimized Technologies to keep Thermoplastics out of Landfills and the Environment (BOTTLE™) Consortium, focusing on materials and recycling innovation. | Continue reading
Integrating symbolic reasoning and learning efficiently from interactions with the world are two major remaining challenges, says vice president and distinguished scientist Nikko Ström. | Continue reading
Meet Amazon Science’s newest research area. | Continue reading
The switch to WebAssembly increases stability, speed. | Continue reading
New voice for Alexa’s Reading Sidekick feature avoids the instabilities common to models with variable prosody. | Continue reading
Team Passing Through, with three academics working together, earns $100,000 first prize. | Continue reading
The story of a decade-plus long journey toward a unified forecasting model. | Continue reading
Submission period opens Aug. 16 and closes on Oct. 8. | Continue reading
Prime Video beats previous state of the art on the MovieNet dataset by 13% with a new model that is 90% smaller and 84% faster. | Continue reading
Dataset includes multiple images of 147,702 products, including 360° rotations and 3-D models for thousands of them. | Continue reading
Alexa Fund company unlocks voice-based computing for people who have trouble using their voices. | Continue reading
Karypis is a featured speaker at the first virtual Amazon Web Services Machine Learning Summit on June 2. | Continue reading
Open-source library enables optimization of hyperparameters to maximize performance while meeting fairness constraints. | Continue reading
Causal analysis improves both the classification accuracy and the relevance of the concepts identified by popular concept-based explanatory models. | Continue reading
Workshop at ICLR 2021 unites communities investigating synthetic data generation to improve machine learning and protect privacy. | Continue reading
Framework improves efficiency, accuracy of applications that search for a handful of solutions in a huge space of candidates. | Continue reading
In a pilot study, an automated code checker found about 100 possible errors, 80% of which turned out to require correction. | Continue reading
Amazon distinguished scientist and conference general chair Alex Smola on what makes MLSys unique — both thematically and culturally. | Continue reading
A model that uses both local and global context improves on the state of the art by 6% and 11% on two benchmark datasets. | Continue reading
The combination of a new loss metric and a module that identifies high-importance image regions improves compression. | Continue reading
How a team of designers, scientists, developers, and engineers worked together to create a truly unique device in Echo Show 10. | Continue reading
Technique that relies on inverse reinforcement learning, or learning by example, improves task completion rate by 14% to 17% in simulations. | Continue reading
Alexa Fund company releases updated and streamlined skill for Alexa that includes "AI Lullaby" soundscape with vocals, music, and voiceovers by Grimes. | Continue reading