The Artificial Intelligence Podcast

New frameworks Google GPipe and Microsoft PipeDream join Uber Horovod in distributed training for deep learning

By pre-processing or post-processing data, or even setting datasets to expire, humans can step in to correct machine learning models

Sharing a container environment and Nvidia GPU server has enabled Domino’s data scientists to create more complex and accurate models to improve store and delivery operations

SQL databases have constraints on data types and consistency. NoSQL does away with them for the sake of speed, flexibility, and scale

Data mining is the automated process of sorting through huge data sets to identify trends and patterns and establish relationships

Like yesterday’s data warehouses, today’s data lakes threaten to lock us into proprietary formats and systems that restrict innovation and raise costs

Analyzing large volumes of data is only part of what makes big data analytics different from traditional data analytics

A data lake can be a much more flexible repository than a data warehouse. Or it can be a trash dump that grows and grows

Intended to ease production deployments of PyTorch models, TorchServe supports multi-model serving and model versioning for A/B testing

TensorFlow, Spark MLlib, Scikit-learn, PyTorch, MXNet, and Keras shine for building and training machine learning and deep learning models