Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.

K-MUG Azure Machine Learning

517 visualizaciones

Publicado el

Slides prepared for K-MUG Azure Machine Learning event on 9 July 2015 at Kochi

Publicado en: Datos y análisis
  • Sé el primero en comentar

  • Sé el primero en recomendar esto

K-MUG Azure Machine Learning

  1. 1. praveen nair blog.ninethsense.com @ninethsense
  2. 2. praveen nair blog.ninethsense.com @ninethsense What is in this presentation? 1. Why Machine Learning? 2. Cortana Analytics Suite 3. High Level view of Azure ML 4. Azure Machine Learning Solution 5. Azure ML Learning Studio Demo 6. Azure ML Market trends Source: All the slides copied form Sales Microsoft Presentations
  3. 3. What is Machine Learning? “Computing systems that improve with experience” Predictive Analytics → “a way to scientifically use the past to predict the future”
  4. 4.  Are you concerned with timely maintenance of in-service equipment?  Are scheduled or routine maintenance operations becoming too costly? Predictive maintenance  Is your current data analysis infrastructure deep enough to support fraud detection?  If not, what are the primary reasons: cost, scale, data science? Frauddetection  Do you currently have social mechanisms in place to promote product recommendations across your site and product portfolio?  If so, are these measures performing well and are you able to analyze customer recommendations for optimal insight? Product recommendation  Do you feel a deeper understanding of historical customer data will help your business predict demand, assess future capacity, or make pricing decisions?  Are you currently using historical data to accurately predict future demand? Demandforecasting  Do you need more accurate data on customer churn and turnover?  Are you losing revenue through customer cancellations? Churnanalysis Sample questionsNeeds Supplychain optimization  Can your organization generate accurate supply chain success metrics, such as Gross Margin Return on Inventory Invested (GMROII) or operational expense minimization?
  5. 5. Azure Machine Learning Microsoft Azure ML → simplifies data analysis and empowers you to find the answers to your business needs A fully managed service that you can use to create, test, operate, and manage predictive analytic solutions in the cloud
  6. 6. Cortana Analytics Suite
  7. 7. Microsoft Azure Machine Learning Data to model to web services in minutes Devices Applications Dashboards Data Storage space Business challenge Business valueModeling Deployment  HDInsight  SQL Server VM  SQL DB  Blobs and tables Cloud Local Clients Integrated development environment for Machine Learning ML Studio http://studio.azureml.net Web API Model is now a web service Microsoft Azure Marketplace Monetize this API Delivering advanced analytics  Desktop files  Excel spreadsheet  Other data files on PC
  8. 8. Deploy in minutes Operationalize models as web services with a single click; monetize in Azure Machine Learning Marketplace Flexible Built-in collection of best of breed algorithms with no coding required. Drop in custom R or use popular CRAN packages Integrated Drag, drop, and connect interface. Data sources with just a drop down run across any data. Fully managed No software to install, no hardware to manage; all you need is an Azure subscription. Built for a cloud-first, mobile-first world Microsoft Azure Machine Learning
  9. 9. The Azure Machine Learning solution Azure ML Studio  Browser-based  Designed for people without deep data science backgrounds  Supports deep science scenarios – R support, multiple models Azure Marketplace  Drag-and-deploy  Fast monetization of ML solutions and APIs  Quick source for free and third-party Azure ML APIs Azure cloud services  No software to install or infrastructure needed  Nearly unlimited file repositories via Azure Storage  Supports Azure data-related services – HDInsight, SQL Database Azure ML API  REST-based web service  Supports best-in-class algorithms  Reduces time from model experimentation to production https://studio.azureml.net/
  10. 10. Write out the results: Azure blob SQL Database Azure table Hive Query (Hadoop) Load Data from…
  11. 11. Demo Time Azure ML Learning Studio
  12. 12. Cortana Intelligence Gallery • https://gallery.cortanaintelligence.com/ • Community of Developers and Data Scientists
  13. 13. T4T3T2T1 CLOUD PLATFORM VENDORS OPEN SOURCE MEGA VENDORS STARTUPS IBM SPSS Python Amazon (Spark ML) Alteryx Google Prediction API R 0xdata SAS SAP (InfiniteInsight) Dell (Statsoft) Oracle Data Miner Predixion bigML RapidMiner GraphLab Skytree Guavus Spark ML Weka Knime FICO Model Builder Salford Matlab Tibco Spotfire
  14. 14. Gartner 2016 Magic Quadrant for Advanced Analytics Platforms
  15. 15. Limitations • Still improving… • Your data is in MSFT Data Centers / Data Privacy • Cloud & Data Safety • Only R & Python supported (currently)
  16. 16. Thank You

×