Download presentation
Presentation is loading. Please wait.
1
What’s New in Azure Machine Learning
6/5/2018 5:53 AM BRK2270 What’s New in Azure Machine Learning Matt Winkler Group Program Manager – Microsoft AI&R © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
2
Agenda Setting Context
6/5/2018 5:53 AM Agenda Setting Context What’s New – Preview features of Azure Machine Learning Getting Started – Building your first project What’s Next? © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
3
A Quick Survey of the Room
4
data science & AI Key trends challenges
Accelerating adoption of AI by developers (consuming models) Rise of hybrid training and scoring scenarios Push scoring/inference to the event (edge, cloud, on-prem) Some developers moving into deep learning as non-traditional path to DS / AI dev Growth of diverse hardware arms race across all form factors (CPU / GPU / FPGA / ASIC / device) Data prep Model deployment & management Model lineage & auditing Explain-ability © 2016 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
5
Drone-based electric grid inspector powered by deep learning
Challenge Traditional power line inspection services are costly Demand for low cost image scoring and support for multiple concurrent customers Needed powerful AI to execute on a drone solution Solution Deep learning to analyze multiple streaming data feeds Azure GPUs support Single Shot multibox detectors Reliable, consistent, and highly elastic scalability with Azure Batch Shipyards
6
Identifying Snow Leopards Computer vision and classification on Spark
Image Image features Decision tree or logistic regression Class 1 Gap snow leopard? Spark ML classifier 34-layer residual image 7x7 conv, 64, /2 pool, /2 3x3 conv, 64 3x3 conv, 64 3x3 conv, 64 3x3 conv, 64 3x3 conv, 64 3x3 conv, 64 3x3 conv, 128, /2 3x3 conv, 128 3x3 conv, 128 3x3 conv, 128 3x3 conv, 128 3x3 conv, 128 3x3 conv, 128 3x3 conv, 128 3x3 conv, 256, /2 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 3x3 conv, 256 Deep neural network Spark ML classifier
7
The AI Development lifecycle
. Social LOB Graph IoT Image CRM Apps + insights INGEST STORE Azure Machine Learning PREP & TRAIN MODEL & SERVE On-prem Cloud Data orchestration and monitoring Data lake and storage Hadoop/Spark/SQL and ML IoT Cloud
8
Azure Machine Learning Studio
Platform for emerging data scientists to graphically build and deploy experiments Rapid experiment composition > 100 easily configured modules for data prep, training, evaluation Extensibility through R & Python Serverless training and deployment Some numbers: 100’s of thousands of deployed models serving billions of requests
9
What have we learned? Customers have told us they love the convenience
Customers have told us they need: Greater control over compute & data More options for model deployment Which frameworks? ALL OF THEM!
10
Key Goals for Preview Features
Build, deploy, and manage models at scale Boost productivity with agile development Begin building now with the tools and platforms you know
11
New Capabilities AZURE MACHINE LEARNING SERVICES
6/5/2018 5:53 AM New Capabilities AZURE MACHINE LEARNING SERVICES Spark SQL Server Virtual machines GPUs Container services Machine Learning Server ON-PREMISES EDGE Azure IoT Edge TRAIN & DEPLOY OPTIONS AZURE Experimentation and Model Management Services Notebooks IDEs Azure Machine Learning Workbench VS Code Tools for AI © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
12
How Do I Get It?
16
What Did I Get?
17
What Did I Get? Experimentation Account Keep track of projects
Local, scale-up, and scale-out training Run history tracking Model Management Account Create containers for models Manage and monitor deployed models
18
Experimentation service
19
Experiment Everywhere
6/5/2018 5:53 AM Experiment Everywhere Local machine Scale up to DSVM AZURE ML EXPERIMENTATION Command line tools IDEs Notebooks in Workbench VS Code Tools for AI Scale out with Spark on HDInsight Azure Batch AI (Coming Soon) ML Server © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
20
Experimentation service
Manage project dependencies Manage training jobs locally, scaled-up or scaled-out Git based checkpointing and version control Service side capture of run metrics, output logs and models Use your favorite IDE, and any framework Use any framework or library Use any tool Use the most popular innovations
21
Model Management service
22
Deploy Everywhere AZURE ML MODEL MANAGEMENT DOCKER
6/5/2018 5:53 AM Deploy Everywhere Single node deployment (cloud/on-prem) Azure Container Service Azure IoT Edge AZURE ML MODEL MANAGEMENT DOCKER Microsoft ML Server Spark clusters SQL Server © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
23
Manage models Deployment and management of models as HTTP services
6/5/2018 5:53 AM Manage models Deployment and management of models as HTTP services Container-based hosting of real time and batch processing Management and monitoring through Azure Application Insights First class support for SparkML, Python, Cognitive Toolkit, TF, R, extensible to support others (Caffe, MXnet) Service authoring in Python © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
24
AI Powered Spreadsheets
Any model published through Model Management can be called from Excel
25
VS Code Tools for AI
26
6/5/2018 5:53 AM VS Code Tools for AI VS Code extension with deep integration to Azure ML End to end development environment, from new project through training Support for remote training Job management On top of all of the goodness of VS Code (Python, Jupyter, Git, etc) © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
27
Workbench
28
What Is It? Windows and Mac based companion for AI development
6/5/2018 5:53 AM What Is It? Windows and Mac based companion for AI development Full environment set up (Python, Jupyter, etc) Embedded notebooks Run History and Comparison experience New data wrangling tools © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
29
AI Powered Data Wrangling
6/5/2018 5:53 AM AI Powered Data Wrangling Rapidly sample, understand, and prep data Leverage PROSE and more for intelligent, data prep by example Extend/customize transforms and featurization through Python Generate Python and Pyspark for execution at scale © 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
30
6/5/2018 5:53 AM Demo © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
31
Demo Summary Project Creation in VS Code Show the code
Job submission locally and to remote GPU Model deployment Run comparison in Workbench Notebooks in Workbench Data preparation in Workbench Sending feedback!
32
Service Pricing Experimentation account – per user (first 2 free)
Model Management account – tiers based on deployment scale (dev/test tier free) Tools – Workbench, VS Code Tools for AI – FREE
33
Machine Learning & AI Portfolio When to use what?
6/5/2018 5:53 AM Machine Learning & AI Portfolio When to use what? Microsoft ML & AI products Build your own or consume pre- trained models? Build your own Consume Azure Machine Learning Cognitive services, bots Which experience do you want? Code first Visual tooling Deployment target (On-prem) ML Server (cloud) AML services (Preview) (cloud) AML Studio What engine(s) do you want to use? On-prem Hadoop SQL Server SQL Server Spark Hadoop Azure Batch DSVM Azure Container Service © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
34
Where can I learn more?
35
Summary Introduction to new Azure ML services Experimentation Service
Model Management Service Introduction to new Azure ML tools VS Code Tools for AI Azure Machine Learning Workbench
36
What’s next Go and get started! Learn more!
6/5/2018 5:53 AM What’s next Go and get started! Learn more! Tell us what you think! Roadmap: Adding support for Azure Batch AI Rolling out to more regions (tell us where) Onward to GA Listening and learning from you © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
37
Ignite 2017 Data Science, ML, and AI Sessions
Session Code Title Speaker Start Time End Time Location BRK3298 How to build machine learning applications using R and Python in SQL Server 2017 Sumit Kumar Umachandar Jayachandran 09/25/ :00 09/25/ :15 Hyatt Regency Windermere W BRK2270 First look at What’s New in Azure Machine Learning Matt Winkler Hyatt Regency Windermere X BRK3301 Let’s talk about Conversation Design Vishwac Sena Kannan Hyatt Plaza International G BRK2268 How to implement the Microsoft Team Data Science Process in data science consulting Debraj GuhaThakurta Brad Johnson 09/26/ :00 09/26/ :15 BRK3299 Language with Microsoft Cognitive Services: infusing language and speech capabilities into your apps Giampaolo Battaglia Luis Cabrera-Cordon Hyatt Plaza International D-F BRK2271 Using big data, the cloud, and AI to enable intelligence at scale Danielle Dean Wee Hyong Tok BRK2269 How to modernize analytics by migrating from SAS to the Microsoft platform Derek Norton Bill Eldredge 09/26/ :45 09/26/ :00 BRK3390 Ship secure faster with Microsoft Security Risk Detection David Molnar BRK3319 Increase the rate of experimentation with Azure Machine Learning Sandhya Vankamamidi Ahmet Gyger Hyatt Plaza International I-K BRK2370 AI development using Data Science Virtual Machines (DSVM) in Azure Barnam Bora Gopi Kumar 09/26/ :30 09/26/ :45 BRK3300 Patterns, Architecture, & Best Practices: Scaling Machine Learning Algorithms with Azure HDInsight Xiaoyong Zhu Hyatt Regency Windermere Z BRK3296 Microsoft Cognitive Services - Infuse intelligence into your business app with Vision and Speech Jennifer Marsman Darren Jefford Hyatt Regency Windermere Y
38
Ignite 2017 Data Science, ML, and AI Sessions
Session Code Title Speaker Start Time End Time Location BRK4033 Deep dive with Microsoft Cognitive Toolkit Cha Zhang 09/26/ :15 09/26/ :30 Hyatt Regency Windermere X BRK4036 Bot Framework patterns and practices straight from our customer Robert Standefer Cindy Noteboom Hyatt Regency Windermere Z BRK2290 Operationalize your models with Azure Machine Learning Chhavi Bhasin Jacob Spoelstra 09/26/ :00 09/26/ :15 Hyatt Plaza International G BRK4034 Working with models for machine learning and Azure Batch AI Alex Sutton Hyatt Plaza International D-F BRK2291 Data Scientists & Developers: Be more productive with Visual Studio and Jupyter Notebooks Shahrokh Mortazavi BRK2289 Azure Revenue Forecasting using Ensemble learning Siddharth Kumar 09/26/ :00 09/26/ :45 Hyatt Plaza International H BRK3334 Building image classification using the Microsoft AI platform Anusua Trivedi Alok Kirpal 09/26/ :15 09/26/ :00 BRK2277 Artificial intelligence: Past, present, but is the future just about machine learning? Rafal Lukawiecki 09/26/ :30 09/26/ :15 Hyatt Regency Windermere W BRK3292 How Jack Henry & Associates is revolutionizing risk prediction & member attrition with credit unions Shau Phang Patty Moore 09/26/ :45 09/26/ :30 BRK3290 Accuracy, reliability, model cross-validation and p-values 09/26/ :00 09/26/ :45 BRK3293 How the Portland Trail Blazers use personalization and Acxiom data to target customers Chris Hoder 09/26/ :15
39
THANK YOU!!!
40
Please evaluate this session
Tech Ready 15 6/5/2018 Please evaluate this session From your Please expand notes window at bottom of slide and read. Then Delete this text box. PC or tablet: visit MyIgnite Phone: download and use the Microsoft Ignite mobile app Your input is important! © 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
41
6/5/2018 5:53 AM © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
Similar presentations
