Data and AI Scientist @ Microsoft
Cloud Solution Architect
US CTO Customer Success
@marktabnet

© Microsoft Corporation
Agenda
• Why Automated Machine Learning?
• Azure ML Service and Azure Databricks
• Capabilities: What’s New?
• Demos
• Enterprise Deployment

Domain specific pretrained models
To simplify solution development
Azure
Databricks
Machine
Learning VMs
Popular frameworks
To build advanced deep learning solutions
TensorFlowPyTorch ONNX
Azure Machine
Learning
LanguageSpeech
…
SearchVision
Productive services
To empower data science and development teams
Powerful infrastructure
To accelerate deep learning
Scikit-Learn
Familiar Data Science tools
To simplify model development
CPU GPU FPGA
From the Intelligent Cloud to the Intelligent Edge
Azure Notebooks JupyterVisual Studio Code Command line

© Microsoft Corporation

Why Automated ML?

Machine Learning on Azure
Domain Specific Pretrained Models
To reduce time to market
Azure
Databricks
Machine
Learning VMs
Popular Frameworks
To build machine learning and deep learning solutions TensorFlowPyTorch ONNX
Azure Machine Learning
LanguageSpeech
…
SearchVision
Productive Services
To empower data science and development teams
Powerful Hardware
To accelerate deep learning
Scikit-Learn
PyCharm Jupyter
Familiar Data Science Tools
To simplify model development Visual Studio Code Command line
CPU GPU FPGA
From the Intelligent Cloud to the Intelligent Edge

Building blocks for a Data Science Project
Data
sources

What is automated machine
learning?
© Microsoft Corporation
Automated machine learning (automated ML) automates
feature engineering, algorithm and hyperparameter
selection to find the best model for your data.

Automated ML Mission
Democratize AI Scale AIAccelerate AI
© Microsoft Corporation Azure
Enable automated building of machine learning with the goal of accelerating, democratizing and scaling AI
Enable Domain Experts & Developers to
get rapidly build AI solutions
Improve Productivity for Data Scientists,
Citizen Data Scientists, App Developers &
Analysts
Build AI solutions at scale in an automated
fashion

How much is this car worth?
Machine Learning Problem Example

Model Creation Is Typically Time-Consuming
Mileage
Condition
Car brand
Year of make
Regulations
…
Parameter 1
Parameter 2
Parameter 3
Parameter 4
…
Gradient Boosted
Nearest Neighbors
SVM
Bayesian Regression
LGBM
…
Mileage Gradient Boosted Criterion
Loss
Min Samples Split
Min Samples Leaf
Others Model
Which algorithm? Which parameters?Which features?
Car brand
Year of make

Criterion
Loss
Min Samples Split
Min Samples Leaf
Others
N Neighbors
Weights
Metric
P
Others
Which algorithm? Which parameters?Which features?
Mileage
Condition
Car brand
Year of make
Regulations
…
Gradient Boosted
Nearest Neighbors
SVM
Bayesian Regression
LGBM
…
Nearest Neighbors
Model
Iterate
Gradient BoostedMileage
Car brand
Year of make
Car brand
Year of make
Condition
Model Creation Is Typically Time-Consuming

Which algorithm? Which parameters?Which features?
Iterate
Model Creation Is Typically Time-Consuming

Enter data
Define goals
Apply constraints
Output
Automated ML Accelerates Model Development
Input Intelligently test multiple models in parallel
Optimized model

Automated ML Capabilities
• Based on Microsoft Research
• Brain trained with several
million experiments
• Collaborative filtering and
Bayesian optimization
• Privacy preserving: No need
to “see” the data

Automated ML Capabilities
• ML Scenarios: Classification &
Regression, Forecasting
• Languages: Python SDK for
deployment and hosting for
inference – Jupyter notebooks
• Training Compute: Local
Machine, AML Compute, Data
Science Virtual Machine (DSVM),
Azure Databricks*
• Transparency: View run history,
model metrics, explainability*
• Scale: Faster model training
using multiple cores and parallel
experiments
* In Preview

Guardrails
Class imbalance
Train-Test split, CV, rolling CV
Missing value imputation
Detect high cardinality features
Detect leaky features
Detect overfitting
Model Interpretability / Feature Importance

About Azure ML Service
and Azure Databricks

+
To empower data science and development teams
Develop models faster with automated machine learning
Use any Python environment and ML frameworks
Manage models across the cloud and the edge.
Prepare data clean data at massive scale
Enable collaboration between data scientists and data engineers
Access machine learning optimized clusters
Azure Machine Learning
Python-based machine learning service
Azure Databricks
Apache Spark-based big-data service

Bring AI to everyone with an end-to-end, scalable, trusted platform
Built with your needs in mind
Support for open source frameworks
Managed compute
DevOps for machine learning
Simple deployment
Tool agnostic Python SDK
Automated machine learning
Seamlessly integrated with the Azure Portfolio
Boost your data science productivity
Increase your rate of experimentation
Deploy and manage your models everywhere

Fast, easy, and collaborative Apache Spark™-based analytics platform
Built with your needs in mind
Optimized Apache Spark environmnet
Collaborative workspace
Integration with Azure data services
Autoscale and autoterminate
Optimized for distributed processing
Support for multiple languages and libraries
Seamlessly integrated with the Azure Portfolio
Increase productivity
Build on a secure, trusted cloud
Scale without limits

Leverage your favorite deep learning frameworks
AZURE ML SERVICE
Increase your rate of experimentation
Bring AI to the edge
Deploy and manage your models everywhere
TensorFlow MS Cognitive Toolkit PyTorch Scikit-Learn ONNX Caffe2 MXNet Chainer
AZURE DATABRICKS
Accelerate processing with the fastest Apache Spark engine
Integrate natively with Azure services
Access enterprise-grade Azure security

What to use when?
+
Customer journey Data Prep Build and Train Manage and Deploy
Apache Spark / Big Data
Python ML developer
Azure ML service
(Pandas, NumPy etc. on AML Compute)
Azure ML service
(OSS frameworks, Hyperdrive, Pipelines,
Automated ML, Model Registry)
Azure ML service
(containerize, deploy,
inference and monitor)
Azure ML service
(containerize, deploy,
inference and monitor)
Azure Databricks
(Apache Spark Dataframes,
Datasets, Delta, Pandas, NumPy etc.)
Azure Databricks + Azure ML service
(Spark MLib and OSS frameworks +
Automated ML, Model Registry)

What’s new?

Latest announcements @ MS Build (Blog post with all the announcements)
Feature engineering updates
• Additional data guardrails and synthetic features
• Added XGBoost algorithm
• Improved transparency retrieving the engineered
features
© Microsoft Corporation Azure
Coming up next
• Improved feature sweeping, text featurization
• Transparency: Get auto-featurized data

Latest announcements @ MS Build (Blog post with all the announcements)
Time Series Forecasting Generally
Available
• Rolling cross validation splits for time series data
• Configurable lags
• Window aggregation
• Holiday featurizer
© Microsoft Corporation Azure
https://azure.microsoft.com/
en-us/blog/build-more-
accurate-forecasts-with-
new-capabilities-in-
automated-machine-
learning/

Latest announcements @ MS Build (Blog post with all the announcements)
Automated ML in ML.NET Model
Builder (Preview)
• Train ML models from Visual Studio
• Inference from your application
© Microsoft Corporation Azure
ML.NET Model Builder

Latest announcements @ MS Build (Blog post with all the announcements)
ONNX support
• Automated ML output ONNX format models
• Inferencing support for C# apps via ONNX runtime
environments (WinML, ML.Net, ONNX C# API), Cosmos
pipelines
© Microsoft Corporation Azure

Latest announcements @ MS Build (Blog post with all the announcements)
Run automated ML from SQL
© Microsoft Corporation Azure
Blog post

Latest announcements @ MS Build (Blog post with all the announcements)
Automated ML UI in Azure portal (Preview)
• End-to-end no-code experience for non-data scientists to
train ML models
• Classification, Regression, Forecasting
• Deploy models easily and quickly
• Advanced settings for power users to tune the training job
© Microsoft Corporation Azure
Blog post Coming up next
• Model explainability
• Additional data sources
(with Datasets)
• Re-run experiments

Demo: Azure Machine
Learning Service

https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-create-portal-experiments

Demo: Azure Databricks

About Azure Databricks
• Azure Databricks is a managed Spark offering on Azure and customers
already use it for advanced analytics.
• It provides a collaborative Notebook based environment with CPU or
GPU based compute cluster.

Azure Databricks Features
• Customers who use Azure Databricks for advanced analytics can now use the
same cluster to run experiments with or without automated machine learning.
• You may keep the data within the same cluster.
• You may leverage the local worker nodes with autoscale and auto termination
capabilities.
• You may use multiple cores of your Azure Databricks cluster to perform
simultaneous training.
• You may further tune the model generated by automated machine learning.
• Every run (including the best run) is available as a pipeline, which you may tune
further if needed.
• The model trained using Azure Databricks can be registered in Azure ML SDK
workspace and then deployed to Azure managed compute (ACI or AKS) using the
Azure Machine learning SDK.

Github Demo
https://github.com/Azure/MachineLearningNotebooks/tree/master/how-to-use-azureml/azure-databricks

How to Configure Azure Databricks
https://docs.microsoft.com/en-us/azure/machine-learning/service/how-to-configure-environment#azure-databricks

Enterprise deployment

Deploy Azure ML models at scale
Azure Machine Learning Service

Model deployment

https://docs.microsoft.com/en-us/azure/architecture/reference-architectures/

Action

Try it for free
http://aka.ms/amlfree
Learn more : https://aka.ms/automatedmldocs
Notebook Samples : https://aka.ms/automatedmlsamples
Blog Post : https://aka.ms/AutomatedML
Product Feedback : AskAutomatedML@microsoft.com