Skip to footer

AI

Omphalos, Uber’s Parallel and Language-Extensible Time Series Backtesting Tool

Uber Engineering created Omphalos, our new backtesting framework, to enable efficient and reliable comparison of forecasting models across languages.

Introducing LCA: Loss Change Allocation for Neural Network Training

Uber AI Labs proposes Loss Change Allocation (LCA), a new method that provides a rich window into the neural network training process.

Faster Neural Networks Straight from JPEG

Uber AI Labs introduces a method for making neural networks that process images faster and more accurately by leveraging JPEG representations.
Photo of Uber app showing map

Applying Customer Feedback: How NLP & Deep Learning Improve Uber’s Maps

To improve our maps, Uber Engineering analyzes customer support tickets with natural language processing and deep learning to identify and correct inaccurate map data.

Gaining Insights in a Simulated Marketplace with Machine Learning at Uber

Uber's Marketplace simulation platform leverages ML to rapidly prototype and test new product features and hypotheses in a risk-free environment.

How to Get a Better GAN (Almost) for Free: Introducing the Metropolis-Hastings GAN

Metropolis-Hastings Generative Adversarial Networks (GANs) leverage the discriminator to pick better samples from the generator after ML model training is done.

Introducing EvoGrad: A Lightweight Library for Gradient-Based Evolution

Uber AI Labs releases EvoGrad, a library for catalyzing gradient-based evolution research, and Evolvability ES, a new meta-learning algorithm enabled by this library.
Complex freeway interchange

Accessible Machine Learning through Data Workflow Management

Uber engineers offer two common use cases showing how we orchestrate machine learning model training in our data workflow engine.

Introducing the Uber Research Publications Site

Uber's Chief Scientist announces the launch of the Uber Research Publications Site, a portal for showcasing our contributions to the research community.

First Uber Science Symposium: Discussing the Next Generation of RL, NLP, ConvAI, and DL

The Uber Science Symposium featured talks from members of the broader scientific community about the the latest innovations in RL, NLP, and other fields.

No Coding Required: Training Models with Ludwig, Uber’s Open Source Deep Learning Toolbox

Uber AI's Piero Molino discusses Ludwig's origin story, common use cases, and how others can get started with this powerful deep learning framework built on top of TensorFlow.
Evolution to running

Accelerating Deep Neuroevolution: Train Atari in Hours on a Single Personal Computer

Applying hardware acceleration to deep neuroevolution in what is now an open source project, Uber AI Labs was able to train a neural network to play Atari in just a few hours on a single personal computer, making this type of research accessible to a far greater number of people.

Evolving Michelangelo Model Representation for Flexibility at Scale

To accommodate additional ML use cases, Uber evolved Michelangelo's application of the Apache Spark MLlib library for greater flexibility and extensibility.

Creating a Zoo of Atari-Playing Agents to Catalyze the Understanding of Deep Reinforcement Learning

Uber AI Labs releases Atari Model Zoo, an open source repository of both trained Atari Learning Environment agents and tools to better understand them.

Three Approaches to Scaling Machine Learning with Uber Seattle Engineering

At an April 2019 meetup on ML and AI at Uber Seattle, members of our engineering team discussed three different approaches to enhancing our ML ecosystem.
Neuron image

Differentiable Plasticity: A New Method for Learning to Learn

Differentiable Plasticity is a new machine learning method for training neural networks to change their connection weights adaptively even after training is completed, allowing a form of learning inspired by the lifelong plasticity of biological brains.

NVIDIA: Accelerating Deep Learning with Uber’s Horovod

Horovod, Uber's open source distributed deep learning system, enables NVIDIA to scale model training from one to eight GPUs for their self-driving sensing and perception technologies.

Ludwig v0.2 Adds New Features and Other Improvements to its Deep Learning Toolbox

Ludwig version 0.2 integrates with Comet.ml, adds a new serving functionality, and incorporates the BERT text encoder, among other new features.

Science at Uber: Applying Artificial Intelligence at Uber

Zoubin Ghahramani, Head of Uber AI, discusses how we use artificial intelligence techniques to make our platform more efficient for users.

Measuring the Intrinsic Dimension of Objective Landscapes

Curious about what it is like to traverse the high-dimensional loss landscapes of modern neural networks? Check out Uber AI Labs’ latest research on measuring intrinsic dimension to find out.

Popular Articles