AWS-SageMaker

• SageMaker Autopilot is the industry’s first automated machine learning capability that gives complete control and visibility into ML models.

• Autopilot automatically inspects raw data, applies feature processors, picks the best set of algorithms, trains & tunes multiple models.

• Users get full visibility into how the model was created and what’s in it & SageMaker Autopilot integrates with SageMaker Studio.

• It  Autopilot can be used by people without machine learning experience to easily produce a model.

• It  Studio provides a single, web-based visual interface where users can perform all ML development steps.

• It Studio gives users complete access, control & visibility into each step required to build, train & deploy models.

• It Autopilot is a generic automatic ML solution for classification and regression problems, such as fraud detection, churn analysis & targeted marketing.

• Users can train models using SageMaker Autopilot and get full access to the models as well as the pipelines that generated the models.

• It Autopilot supports 2 built-in algorithms at launch: XGBoost and Linear Learner.

• It Autopilot built-in algorithms support distributed training out of the box.

• It Notebooks provide one-click Jupyter notebooks that users can start working with in seconds.

• With SageMaker Notebooks users can sign in with their corporate credentials using SSO and start working with notebooks within seconds.

• It Notebooks give users access to all SageMaker features, such as distributed training, batch transform, hosting & experiment management.

• It Ground Truth provides automated data labeling using machine learning.

• It Ground Truth will first select a random sample of data and send it to Mechanical Turk to be labeled.

• It Experiments helps users organize and track iterations to machine learning models.

• It Experiments helps users manage iterations by automatically capturing the input parameters, configurations and results, and storing them as experiments.

• It Debugger makes the training process more transparent by automatically capturing real-time metrics during training such as training and validation, confusion matrices & learning gradients to help improve model accuracy.

• The metrics from SageMaker Debugger can be visualized in SageMaker Studio for easy understanding.

• It Debugger can also generate warnings and remediation advice when common training problems are detected.

• It is a fully-managed service that enables data scientists and developers to quickly and easily build, train & deploy machine learning models.

• It enables developers and scientists to build machine learning models for use in intelligent, predictive apps.

• It is designed for high availability. There are no maintenance windows or scheduled downtimes.

• It APIs run in Amazon’s proven, high-availability data centers, with service stack replication configured across three facilities in each AWS region to provide fault tolerance in the event of a server failure or AZ outage.

• It ensures that ML model artifacts and other system artifacts are encrypted in transit and at rest.

• Requests to the SageMaker API and console are made over a secure (SSL) connection.

• It stores code in ML storage volumes, secured by security groups and optionally encrypted at rest.

• It allows users to select the number and type of instance used for the hosted notebook, training & model hosting.

• It provides a full end-to-end workflow, but users can continue to use their existing tools with it.

• Users pay for ML compute, storage and data processing resources their use for hosting the notebook, training the model, performing predictions & logging the outputs.

• It supports Jupyter notebooks.

• Users can persist their notebook files on the attached ML storage volume.

• Users can modify the notebook instance and select a larger profile through the SageMaker console, after saving their files and data on the attached ML storage volume.

• Managed Spot Training with SageMaker lets users train their machine learning models using EC2 Spot instances, while reducing the cost of training their models by up to 90%.

• Managed Spot Training is supported on all AWS regions where Amazon SageMaker is currently available.

• There are no fixed limits to the size of the dataset users can use for training models with Amazon SageMaker.

• It includes built-in algorithms for linear regression, logistic regression, k-means clustering, principal component analysis, factorization machines, neural topic modeling, latent dirichlet allocation, gradient boosted trees, sequence2sequence, time series forecasting, word2vec & image classification.

• It also provides optimized Apache MXNet, Tensorflow, Chainer & PyTorch containers.

• It supports users custom training algorithms provided through a Docker image adhering to the documented specification.

• User can train reinforcement learning models in SageMaker in addition to supervised and unsupervised learning models.

• It RL supports a number of different environments for training reinforcement learning models.

• It RL includes RL toolkits such as Coach and Ray RLLib that offer implementations of RL agent algorithms such as DQN, PPO, A3C & many more.

• Users can bring their own RL libraries and algorithm implementations in Docker Containers and run those in SageMaker RL.

• It Neo is a new capability that enables machine learning models to train once and run anywhere in the cloud and at the edge.

• It Neo contains two major components – a compiler and a runtime.