SlideShare une entreprise Scribd logo
1  sur  37
Previously known as
Think Big. Move Fast.
Iris Multi-Class Classifier
with Azure ML
A friendly tutorial in 10 Steps
V 1.0 – February 2015
Davide Mauri
dmauri@solidq.com
@mauridb
About this tutorial
• The objective of this tutorial is to show how to use the original IRIS Dataset
with AzureML for multiclass classification
• On AzureML the existing dataset is limited to a binary classification…
• …but the original one is much more interesting!
• We’ll publish the trained model as a web service to be used in your
applications
• You’ll need a
• AzureML Account (free in Preview)
• We’ll use a “Supervised Learning” algorithm
• Specifically the we’ll be using a Neural Network
Useful References
• AzureML
• https://studio.azureml.net/
• Machine Learning
• http://en.wikipedia.org/wiki/Machine_learning
• Supervised Learning
• http://en.wikipedia.org/wiki/Supervised_learning
• Neural Networks
• http://en.wikipedia.org/wiki/Artificial_neural_network
IRIS Dataset
• Available from UC Irvine Machine Learning Repository
• http://archive.ics.uci.edu/ml/datasets/Iris
• Classification of three Iris species, with four features
• Sepal Width & Length, Petal Width & Length
• http://en.wikipedia.org/wiki/Iris_flower_data_set
• «This is perhaps the best known database to be found in the pattern
recognition literature. […] The data set contains 3 classes of 50 instances each,
where each class refers to a type of iris plant. One class is linearly separable
from the other 2; the latter are NOT linearly separable from each other.»
UC Irvine Machine Learning Repository
IRIS Dataset
http://www.anselm.edu/homepage/jpitocch/genbi101/diversity3Plants.html
Create The Dataset
Step 1 – Create Dataset
• Create a New Dataset, by uploading the downloaded IRIS Dataset, and name it
«Iris UCI Dataset»
Train the Model
Step 2 – Create Experiment
• Create a new Blank Experiment
Step 3 – Add Dataset
• Now look for the Iris UCI Dataset in the «Dataset» menu item on the left
• Or search it in the top-left search box
• Drag and drop it in the design area
Step 4 – Add Initialization
• Now the model being developed needs to be initialized and trained.
• Under the Machine Learning menu look for Initialize Model  Classification 
Multiclass Neural Network and drop it on the design area
Step 5 – Split Data
• In order to train the model, only a part of the original dataset will be used. The
remaining part will be used to evaluate it.
• Drop the Split component from the Data Transformation  Sample and Split
menu and connect the Iris UCI Dataset to the Split Input
Step 5 – Split Data
• Usually the dataset is split between 70% for training and 30% for evaluating
the model. Such configuration can be set in the “Properties” pane.
Step 6 – Train Model
• Now the Train Model component has to be added in the design area in order
to have training working. Drop the component from the Train  Train Model
element under Machine Learning menu, and connect the item as shown in
figure:
Step 6 – Train Model
• Now select which column will be used as predictors. The Iris Dataset has 5
columns
• sepal length in cm
• sepal width in cm
• petal length in cm
• petal width in cm
• class
• The class column contains the value that we want to be predicted based on
the value of the others four columns
Step 6 – Train Model
• Select the Train Model component that has been placed on the design area
before and click on the «Launch columns selector» in the options area and
then select the class column
Step 6 – Train Model
• It’s now possible to run the experiment in order to have the model trained.
• Just click on the RUN icon in toolbar to run it
• If everything runs correctly a green thick mark will be placed in every
component:
Step 7 – Save the Trained Model
• The model now can be saved as Trained so that it can be used later.
• Right click on the Train Model component’s output endpoint
• Select «Save as Trained Model»
• Name it NN Test
Score the Model
Step 8 – Score the Model
• Create a new experiment. In the design area add
• Iris UCI Dataset
• Split component
• NN Test from Trained Models
• Score Model from Machine Learning  Score
• Connect
• Iris UCI Dataset to Split
• Second Split output to second Score Model input
• Configure the Split as did before (70/30)
• NN Test to first Score Model input
Step 9 – Score the Model
• Here’s how the resulting experiment should look like:
Step 9 – Score the Model
• Now you can Run the experiment. This time the trained model will be used to
predict the 30% of the data we already know the classification but that wasn’t
used in training.
• That explain the name of «Supervised Learning». We are teaching the model what we
already know so that it can learn how to classify unknown things for use in future
• Once the experiment has finished, you can visualize the scored results, by right
clicking on the Score Model output and select Visualize
Step 9 – Score the Model
• In the Visualize window, select the class column and in the «Visualization»
pane, in the compare to dropdown, select «Scored Labels»
In this case, only 3 out of 45
rows as been wrongly
classified! The success rate is
above 90%!
Publish Web Service
Step 10 - Publish Web Service
• Create a new experiment. In the design area add
• Iris UCI Dataset
• Project Columns from Data Transformation  Manipulation
• NN Test from Trained Models
• Score Model from Machine Learning  Score
Step 10 - Publish Web Service
• The Project Columns will be used to strip the class column from the data
source
• Since the model will predict it
• It will also define the correct metadata when the model will be published as a Web
Service
• Connect it with Iris UCI Dataset and with the Score Model
• Make sure al but class column are selected in Project Columns properties (use column
selector)
• Connect the NN Test to the Score Model
Step 10 - Publish Web Service
• The resulting experiment should look like this
Step 10 - Publish Web Service
• Run the Experiment
• After experiment has finished correctly, add another Project Columns
connected to the Score Model
• In the column selector select «All Scores»
• (At present time) The «All Scores» option is available ONLY after the first run.
• This is need for the web service, to strip out all the source columns and keep only the
results
• Run the experiment again
Step 10 - Publish Web Service
• The resulting experiment should look like this
Step 10 - Publish Web Service
• In order to have the Web Service publish, Input and Output data must be
defined.
• The input will be defined using the metadata of the second Score Model input
• Right click on the input and the select «Set as Publish Input»
Step 10 - Publish Web Service
• The output will be defined using the metadata of the output of the second
Project Model component
• Right click on the output and the select «Set as Publish Output»
Step 10 - Publish Web Service
• The resulting experiment should look like this
Step 10 - Publish Web Service
• Click on the «Publish Web Service» icon
• Now the web service can be tested and give sepal and petal data as input, it
will return the probability for each class and the most probable class as result
• You’ll find the Web Service in the «Web Service» section of AzureML
homepage.
• Web Service also provides a testing page and examples to use it with
• C#, R, Python
Conclusion
Conclusion
• We created three experiments just for the tutorial purpose but actually only
two are needed.
• Experiment one and two (Training and Scoring) can be merged together
• We only used the Neural Network Classifier but there other Multiclass
classifier that could (and should) be used
• Test all of them and take the one that give best predictions
Previously known as
Think Big. Move Fast.

Contenu connexe

Tendances

No Container: a Modern Java Stack with Bootique
No Container: a Modern Java Stack with BootiqueNo Container: a Modern Java Stack with Bootique
No Container: a Modern Java Stack with BootiqueAndrus Adamchik
 
Azure SQL & SQL Server 2016 JSON
Azure SQL & SQL Server 2016 JSONAzure SQL & SQL Server 2016 JSON
Azure SQL & SQL Server 2016 JSONDavide Mauri
 
La vita nella corsia di sorpasso; A tutta velocità, XPages!
La vita nella corsia di sorpasso; A tutta velocità, XPages!La vita nella corsia di sorpasso; A tutta velocità, XPages!
La vita nella corsia di sorpasso; A tutta velocità, XPages!Ulrich Krause
 
Developing Complex WordPress Sites without Fear of Failure (with MVC)
Developing Complex WordPress Sites without Fear of Failure (with MVC)Developing Complex WordPress Sites without Fear of Failure (with MVC)
Developing Complex WordPress Sites without Fear of Failure (with MVC)Mike Schinkel
 
EPUB for Website Producers
EPUB for Website ProducersEPUB for Website Producers
EPUB for Website Producerswaldoj
 
Granting Oracle Schema Permissions when Objects not created Yet
Granting Oracle Schema Permissions when Objects not created YetGranting Oracle Schema Permissions when Objects not created Yet
Granting Oracle Schema Permissions when Objects not created YetSecure-24
 
Marty, You're Just Not Thinking Fourth Dimensionally
Marty, You're Just Not Thinking Fourth DimensionallyMarty, You're Just Not Thinking Fourth Dimensionally
Marty, You're Just Not Thinking Fourth DimensionallyTeamstudio
 
PowerShell for the Anxious ITPro
PowerShell for the Anxious ITProPowerShell for the Anxious ITPro
PowerShell for the Anxious ITProJason Himmelstein
 
Heavy Metal PowerPivot Remastered
Heavy Metal PowerPivot RemasteredHeavy Metal PowerPivot Remastered
Heavy Metal PowerPivot RemasteredJason Himmelstein
 
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...SPTechCon
 
Free Oracle Training
Free Oracle TrainingFree Oracle Training
Free Oracle Trainingsdturton
 
WordPress as a CMS - Case Study of an Organizational Intranet
WordPress as a CMS - Case Study of an Organizational IntranetWordPress as a CMS - Case Study of an Organizational Intranet
WordPress as a CMS - Case Study of an Organizational IntranetTech Liminal
 
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010Corey Roth
 
Introduction to styling Molly
Introduction to styling MollyIntroduction to styling Molly
Introduction to styling MollyChris Northwood
 
SharePoint 2013 Sandbox Solutions for On Premise and Office 365
SharePoint 2013 Sandbox Solutions for On Premise and Office 365SharePoint 2013 Sandbox Solutions for On Premise and Office 365
SharePoint 2013 Sandbox Solutions for On Premise and Office 365Ed Musters
 
A Beginner's Guide to Ember
A Beginner's Guide to EmberA Beginner's Guide to Ember
A Beginner's Guide to EmberRichard Martin
 
Back from the Dead: When Bad Code Kills a Good Server
Back from the Dead: When Bad Code Kills a Good ServerBack from the Dead: When Bad Code Kills a Good Server
Back from the Dead: When Bad Code Kills a Good ServerTeamstudio
 

Tendances (20)

No Container: a Modern Java Stack with Bootique
No Container: a Modern Java Stack with BootiqueNo Container: a Modern Java Stack with Bootique
No Container: a Modern Java Stack with Bootique
 
Azure SQL & SQL Server 2016 JSON
Azure SQL & SQL Server 2016 JSONAzure SQL & SQL Server 2016 JSON
Azure SQL & SQL Server 2016 JSON
 
La vita nella corsia di sorpasso; A tutta velocità, XPages!
La vita nella corsia di sorpasso; A tutta velocità, XPages!La vita nella corsia di sorpasso; A tutta velocità, XPages!
La vita nella corsia di sorpasso; A tutta velocità, XPages!
 
Developing Complex WordPress Sites without Fear of Failure (with MVC)
Developing Complex WordPress Sites without Fear of Failure (with MVC)Developing Complex WordPress Sites without Fear of Failure (with MVC)
Developing Complex WordPress Sites without Fear of Failure (with MVC)
 
Solr
SolrSolr
Solr
 
EPUB for Website Producers
EPUB for Website ProducersEPUB for Website Producers
EPUB for Website Producers
 
Granting Oracle Schema Permissions when Objects not created Yet
Granting Oracle Schema Permissions when Objects not created YetGranting Oracle Schema Permissions when Objects not created Yet
Granting Oracle Schema Permissions when Objects not created Yet
 
Marty, You're Just Not Thinking Fourth Dimensionally
Marty, You're Just Not Thinking Fourth DimensionallyMarty, You're Just Not Thinking Fourth Dimensionally
Marty, You're Just Not Thinking Fourth Dimensionally
 
Philly TechFest SQL Indexes
Philly TechFest SQL IndexesPhilly TechFest SQL Indexes
Philly TechFest SQL Indexes
 
PowerShell for the Anxious ITPro
PowerShell for the Anxious ITProPowerShell for the Anxious ITPro
PowerShell for the Anxious ITPro
 
Heavy Metal PowerPivot Remastered
Heavy Metal PowerPivot RemasteredHeavy Metal PowerPivot Remastered
Heavy Metal PowerPivot Remastered
 
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...
Part I: SharePoint 2013 Administration by Todd Klindt and Shane Young - SPTec...
 
SharePoint Performance
SharePoint PerformanceSharePoint Performance
SharePoint Performance
 
Free Oracle Training
Free Oracle TrainingFree Oracle Training
Free Oracle Training
 
WordPress as a CMS - Case Study of an Organizational Intranet
WordPress as a CMS - Case Study of an Organizational IntranetWordPress as a CMS - Case Study of an Organizational Intranet
WordPress as a CMS - Case Study of an Organizational Intranet
 
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010
SharePoint 2010 Development for ASP.NET Developers - Tyson Devcon 2010
 
Introduction to styling Molly
Introduction to styling MollyIntroduction to styling Molly
Introduction to styling Molly
 
SharePoint 2013 Sandbox Solutions for On Premise and Office 365
SharePoint 2013 Sandbox Solutions for On Premise and Office 365SharePoint 2013 Sandbox Solutions for On Premise and Office 365
SharePoint 2013 Sandbox Solutions for On Premise and Office 365
 
A Beginner's Guide to Ember
A Beginner's Guide to EmberA Beginner's Guide to Ember
A Beginner's Guide to Ember
 
Back from the Dead: When Bad Code Kills a Good Server
Back from the Dead: When Bad Code Kills a Good ServerBack from the Dead: When Bad Code Kills a Good Server
Back from the Dead: When Bad Code Kills a Good Server
 

En vedette

Apache Flink: API, runtime, and project roadmap
Apache Flink: API, runtime, and project roadmapApache Flink: API, runtime, and project roadmap
Apache Flink: API, runtime, and project roadmapKostas Tzoumas
 
Datarace: IoT e Big Data (Italian)
Datarace: IoT e Big Data (Italian)Datarace: IoT e Big Data (Italian)
Datarace: IoT e Big Data (Italian)Davide Mauri
 
Azure Machine Learning
Azure Machine LearningAzure Machine Learning
Azure Machine LearningDavide Mauri
 
AzureML - Creating and Using Machine Learning Solutions (Italian)
AzureML - Creating and Using Machine Learning Solutions (Italian)AzureML - Creating and Using Machine Learning Solutions (Italian)
AzureML - Creating and Using Machine Learning Solutions (Italian)Davide Mauri
 
Real Time Power BI
Real Time Power BIReal Time Power BI
Real Time Power BIDavide Mauri
 
SQL Server 2016 What's New For Developers
SQL Server 2016  What's New For DevelopersSQL Server 2016  What's New For Developers
SQL Server 2016 What's New For DevelopersDavide Mauri
 
Agile Data Warehousing
Agile Data WarehousingAgile Data Warehousing
Agile Data WarehousingDavide Mauri
 
Dashboarding with Microsoft: Datazen & Power BI
Dashboarding with Microsoft: Datazen & Power BIDashboarding with Microsoft: Datazen & Power BI
Dashboarding with Microsoft: Datazen & Power BIDavide Mauri
 
Azure Machine Learning (Italian)
Azure Machine Learning (Italian)Azure Machine Learning (Italian)
Azure Machine Learning (Italian)Davide Mauri
 
Temporal Snapshot Fact Tables
Temporal Snapshot Fact TablesTemporal Snapshot Fact Tables
Temporal Snapshot Fact TablesDavide Mauri
 
Machine learning @ Spotify - Madison Big Data Meetup
Machine learning @ Spotify - Madison Big Data MeetupMachine learning @ Spotify - Madison Big Data Meetup
Machine learning @ Spotify - Madison Big Data MeetupAndy Sloane
 
Modern SQL in Open Source and Commercial Databases
Modern SQL in Open Source and Commercial DatabasesModern SQL in Open Source and Commercial Databases
Modern SQL in Open Source and Commercial DatabasesMarkus Winand
 
An Introduction to Supervised Machine Learning and Pattern Classification: Th...
An Introduction to Supervised Machine Learning and Pattern Classification: Th...An Introduction to Supervised Machine Learning and Pattern Classification: Th...
An Introduction to Supervised Machine Learning and Pattern Classification: Th...Sebastian Raschka
 

En vedette (14)

Apache Flink: API, runtime, and project roadmap
Apache Flink: API, runtime, and project roadmapApache Flink: API, runtime, and project roadmap
Apache Flink: API, runtime, and project roadmap
 
Datarace: IoT e Big Data (Italian)
Datarace: IoT e Big Data (Italian)Datarace: IoT e Big Data (Italian)
Datarace: IoT e Big Data (Italian)
 
Azure Machine Learning
Azure Machine LearningAzure Machine Learning
Azure Machine Learning
 
Data modeling facts
Data modeling factsData modeling facts
Data modeling facts
 
AzureML - Creating and Using Machine Learning Solutions (Italian)
AzureML - Creating and Using Machine Learning Solutions (Italian)AzureML - Creating and Using Machine Learning Solutions (Italian)
AzureML - Creating and Using Machine Learning Solutions (Italian)
 
Real Time Power BI
Real Time Power BIReal Time Power BI
Real Time Power BI
 
SQL Server 2016 What's New For Developers
SQL Server 2016  What's New For DevelopersSQL Server 2016  What's New For Developers
SQL Server 2016 What's New For Developers
 
Agile Data Warehousing
Agile Data WarehousingAgile Data Warehousing
Agile Data Warehousing
 
Dashboarding with Microsoft: Datazen & Power BI
Dashboarding with Microsoft: Datazen & Power BIDashboarding with Microsoft: Datazen & Power BI
Dashboarding with Microsoft: Datazen & Power BI
 
Azure Machine Learning (Italian)
Azure Machine Learning (Italian)Azure Machine Learning (Italian)
Azure Machine Learning (Italian)
 
Temporal Snapshot Fact Tables
Temporal Snapshot Fact TablesTemporal Snapshot Fact Tables
Temporal Snapshot Fact Tables
 
Machine learning @ Spotify - Madison Big Data Meetup
Machine learning @ Spotify - Madison Big Data MeetupMachine learning @ Spotify - Madison Big Data Meetup
Machine learning @ Spotify - Madison Big Data Meetup
 
Modern SQL in Open Source and Commercial Databases
Modern SQL in Open Source and Commercial DatabasesModern SQL in Open Source and Commercial Databases
Modern SQL in Open Source and Commercial Databases
 
An Introduction to Supervised Machine Learning and Pattern Classification: Th...
An Introduction to Supervised Machine Learning and Pattern Classification: Th...An Introduction to Supervised Machine Learning and Pattern Classification: Th...
An Introduction to Supervised Machine Learning and Pattern Classification: Th...
 

Similaire à Iris Multi-Class Classifier with Azure ML

Azure machine learning
Azure machine learningAzure machine learning
Azure machine learningSimone Caldaro
 
Azure Machine Learning Dotnet Campus 2015
Azure Machine Learning Dotnet Campus 2015 Azure Machine Learning Dotnet Campus 2015
Azure Machine Learning Dotnet Campus 2015 antimo musone
 
Build a Sentiment Model using ML.Net
Build a Sentiment Model using ML.NetBuild a Sentiment Model using ML.Net
Build a Sentiment Model using ML.NetCheah Eng Soon
 
Machine learning and azure ml studio
Machine learning and azure ml studioMachine learning and azure ml studio
Machine learning and azure ml studioYogendra Tamang
 
Net campus2015 antimomusone
Net campus2015 antimomusoneNet campus2015 antimomusone
Net campus2015 antimomusoneDotNetCampus
 
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATA
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATAPREDICT THE FUTURE , MACHINE LEARNING & BIG DATA
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATADotNetCampus
 
Quality Center
Quality CenterQuality Center
Quality CenterSachin-QA
 
Test Management Tool HP ALM- Quality Center Part 1
Test Management Tool  HP ALM- Quality Center Part 1Test Management Tool  HP ALM- Quality Center Part 1
Test Management Tool HP ALM- Quality Center Part 1Murageppa-QA
 
HP Quality Center
HP Quality CenterHP Quality Center
HP Quality CenterANKUR-BA
 
Test Management Tool HP ALM- Quality Center Part 2
Test Management Tool  HP ALM- Quality Center Part 2Test Management Tool  HP ALM- Quality Center Part 2
Test Management Tool HP ALM- Quality Center Part 2Murageppa-QA
 
Santander customer satisfaction
Santander customer satisfactionSantander customer satisfaction
Santander customer satisfactionAprameya Bhol
 
PeopleSoft 9.2 HCM Features and Functions Including Fluid Mobile
PeopleSoft 9.2 HCM Features and Functions Including Fluid MobilePeopleSoft 9.2 HCM Features and Functions Including Fluid Mobile
PeopleSoft 9.2 HCM Features and Functions Including Fluid MobileNERUG
 
Concurrency Programming in Java - 02 - Essentials of Java Part 1
Concurrency Programming in Java - 02 - Essentials of Java Part 1Concurrency Programming in Java - 02 - Essentials of Java Part 1
Concurrency Programming in Java - 02 - Essentials of Java Part 1Sachintha Gunasena
 
Cutting Edge Computer Vision for Everyone
Cutting Edge Computer Vision for EveryoneCutting Edge Computer Vision for Everyone
Cutting Edge Computer Vision for EveryoneIvo Andreev
 
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and Tricks
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and TricksIBM Cognos 10 Framework Manager Metadata Modeling: Tips and Tricks
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and TricksSenturus
 
Apache Spark Model Deployment
Apache Spark Model Deployment Apache Spark Model Deployment
Apache Spark Model Deployment Databricks
 
Running with Elephants: Predictive Analytics with HDInsight
Running with Elephants: Predictive Analytics with HDInsightRunning with Elephants: Predictive Analytics with HDInsight
Running with Elephants: Predictive Analytics with HDInsightChris Price
 
Creating Your First Predictive Model In Python
Creating Your First Predictive Model In PythonCreating Your First Predictive Model In Python
Creating Your First Predictive Model In PythonRobert Dempsey
 
A survey on Machine Learning In Production (July 2018)
A survey on Machine Learning In Production (July 2018)A survey on Machine Learning In Production (July 2018)
A survey on Machine Learning In Production (July 2018)Arnab Biswas
 

Similaire à Iris Multi-Class Classifier with Azure ML (20)

Azure machine learning
Azure machine learningAzure machine learning
Azure machine learning
 
Azure Machine Learning Dotnet Campus 2015
Azure Machine Learning Dotnet Campus 2015 Azure Machine Learning Dotnet Campus 2015
Azure Machine Learning Dotnet Campus 2015
 
Build a Sentiment Model using ML.Net
Build a Sentiment Model using ML.NetBuild a Sentiment Model using ML.Net
Build a Sentiment Model using ML.Net
 
Machine learning and azure ml studio
Machine learning and azure ml studioMachine learning and azure ml studio
Machine learning and azure ml studio
 
Net campus2015 antimomusone
Net campus2015 antimomusoneNet campus2015 antimomusone
Net campus2015 antimomusone
 
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATA
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATAPREDICT THE FUTURE , MACHINE LEARNING & BIG DATA
PREDICT THE FUTURE , MACHINE LEARNING & BIG DATA
 
Quality Center
Quality CenterQuality Center
Quality Center
 
Test Management Tool HP ALM- Quality Center Part 1
Test Management Tool  HP ALM- Quality Center Part 1Test Management Tool  HP ALM- Quality Center Part 1
Test Management Tool HP ALM- Quality Center Part 1
 
HP Quality Center
HP Quality CenterHP Quality Center
HP Quality Center
 
Test Management Tool HP ALM- Quality Center Part 2
Test Management Tool  HP ALM- Quality Center Part 2Test Management Tool  HP ALM- Quality Center Part 2
Test Management Tool HP ALM- Quality Center Part 2
 
Santander customer satisfaction
Santander customer satisfactionSantander customer satisfaction
Santander customer satisfaction
 
PeopleSoft 9.2 HCM Features and Functions Including Fluid Mobile
PeopleSoft 9.2 HCM Features and Functions Including Fluid MobilePeopleSoft 9.2 HCM Features and Functions Including Fluid Mobile
PeopleSoft 9.2 HCM Features and Functions Including Fluid Mobile
 
Concurrency Programming in Java - 02 - Essentials of Java Part 1
Concurrency Programming in Java - 02 - Essentials of Java Part 1Concurrency Programming in Java - 02 - Essentials of Java Part 1
Concurrency Programming in Java - 02 - Essentials of Java Part 1
 
Model builder in_arcgis
Model builder in_arcgisModel builder in_arcgis
Model builder in_arcgis
 
Cutting Edge Computer Vision for Everyone
Cutting Edge Computer Vision for EveryoneCutting Edge Computer Vision for Everyone
Cutting Edge Computer Vision for Everyone
 
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and Tricks
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and TricksIBM Cognos 10 Framework Manager Metadata Modeling: Tips and Tricks
IBM Cognos 10 Framework Manager Metadata Modeling: Tips and Tricks
 
Apache Spark Model Deployment
Apache Spark Model Deployment Apache Spark Model Deployment
Apache Spark Model Deployment
 
Running with Elephants: Predictive Analytics with HDInsight
Running with Elephants: Predictive Analytics with HDInsightRunning with Elephants: Predictive Analytics with HDInsight
Running with Elephants: Predictive Analytics with HDInsight
 
Creating Your First Predictive Model In Python
Creating Your First Predictive Model In PythonCreating Your First Predictive Model In Python
Creating Your First Predictive Model In Python
 
A survey on Machine Learning In Production (July 2018)
A survey on Machine Learning In Production (July 2018)A survey on Machine Learning In Production (July 2018)
A survey on Machine Learning In Production (July 2018)
 

Plus de Davide Mauri

Azure serverless Full-Stack kickstart
Azure serverless Full-Stack kickstartAzure serverless Full-Stack kickstart
Azure serverless Full-Stack kickstartDavide Mauri
 
Agile Data Warehousing
Agile Data WarehousingAgile Data Warehousing
Agile Data WarehousingDavide Mauri
 
Dapper: the microORM that will change your life
Dapper: the microORM that will change your lifeDapper: the microORM that will change your life
Dapper: the microORM that will change your lifeDavide Mauri
 
When indexes are not enough
When indexes are not enoughWhen indexes are not enough
When indexes are not enoughDavide Mauri
 
Building a Real-Time IoT monitoring application with Azure
Building a Real-Time IoT monitoring application with AzureBuilding a Real-Time IoT monitoring application with Azure
Building a Real-Time IoT monitoring application with AzureDavide Mauri
 
SSIS Monitoring Deep Dive
SSIS Monitoring Deep DiveSSIS Monitoring Deep Dive
SSIS Monitoring Deep DiveDavide Mauri
 
Azure Stream Analytics
Azure Stream AnalyticsAzure Stream Analytics
Azure Stream AnalyticsDavide Mauri
 
Event Hub & Azure Stream Analytics
Event Hub & Azure Stream AnalyticsEvent Hub & Azure Stream Analytics
Event Hub & Azure Stream AnalyticsDavide Mauri
 
SQL Server 2016 JSON
SQL Server 2016 JSONSQL Server 2016 JSON
SQL Server 2016 JSONDavide Mauri
 
Back to the roots - SQL Server Indexing
Back to the roots - SQL Server IndexingBack to the roots - SQL Server Indexing
Back to the roots - SQL Server IndexingDavide Mauri
 
BIML: BI to the next level
BIML: BI to the next levelBIML: BI to the next level
BIML: BI to the next levelDavide Mauri
 
Data Science Overview
Data Science OverviewData Science Overview
Data Science OverviewDavide Mauri
 
Delayed durability
Delayed durabilityDelayed durability
Delayed durabilityDavide Mauri
 
Hekaton: In-memory tables
Hekaton: In-memory tablesHekaton: In-memory tables
Hekaton: In-memory tablesDavide Mauri
 
Hardware planning & sizing for sql server
Hardware planning & sizing for sql serverHardware planning & sizing for sql server
Hardware planning & sizing for sql serverDavide Mauri
 

Plus de Davide Mauri (16)

Azure serverless Full-Stack kickstart
Azure serverless Full-Stack kickstartAzure serverless Full-Stack kickstart
Azure serverless Full-Stack kickstart
 
Agile Data Warehousing
Agile Data WarehousingAgile Data Warehousing
Agile Data Warehousing
 
Dapper: the microORM that will change your life
Dapper: the microORM that will change your lifeDapper: the microORM that will change your life
Dapper: the microORM that will change your life
 
When indexes are not enough
When indexes are not enoughWhen indexes are not enough
When indexes are not enough
 
Building a Real-Time IoT monitoring application with Azure
Building a Real-Time IoT monitoring application with AzureBuilding a Real-Time IoT monitoring application with Azure
Building a Real-Time IoT monitoring application with Azure
 
SSIS Monitoring Deep Dive
SSIS Monitoring Deep DiveSSIS Monitoring Deep Dive
SSIS Monitoring Deep Dive
 
Azure Stream Analytics
Azure Stream AnalyticsAzure Stream Analytics
Azure Stream Analytics
 
Event Hub & Azure Stream Analytics
Event Hub & Azure Stream AnalyticsEvent Hub & Azure Stream Analytics
Event Hub & Azure Stream Analytics
 
SQL Server 2016 JSON
SQL Server 2016 JSONSQL Server 2016 JSON
SQL Server 2016 JSON
 
Back to the roots - SQL Server Indexing
Back to the roots - SQL Server IndexingBack to the roots - SQL Server Indexing
Back to the roots - SQL Server Indexing
 
BIML: BI to the next level
BIML: BI to the next levelBIML: BI to the next level
BIML: BI to the next level
 
Data juice
Data juiceData juice
Data juice
 
Data Science Overview
Data Science OverviewData Science Overview
Data Science Overview
 
Delayed durability
Delayed durabilityDelayed durability
Delayed durability
 
Hekaton: In-memory tables
Hekaton: In-memory tablesHekaton: In-memory tables
Hekaton: In-memory tables
 
Hardware planning & sizing for sql server
Hardware planning & sizing for sql serverHardware planning & sizing for sql server
Hardware planning & sizing for sql server
 

Dernier

Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Delhi Call girls
 
Edukaciniai dropshipping via API with DroFx
Edukaciniai dropshipping via API with DroFxEdukaciniai dropshipping via API with DroFx
Edukaciniai dropshipping via API with DroFxolyaivanovalion
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysismanisha194592
 
CebaBaby dropshipping via API with DroFX.pptx
CebaBaby dropshipping via API with DroFX.pptxCebaBaby dropshipping via API with DroFX.pptx
CebaBaby dropshipping via API with DroFX.pptxolyaivanovalion
 
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Online
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service OnlineCALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Online
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Onlineanilsa9823
 
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 night
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 nightCheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 night
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 nightDelhi Call girls
 
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...shivangimorya083
 
Invezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz1
 
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAl Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAroojKhan71
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxfirstjob4
 
Halmar dropshipping via API with DroFx
Halmar  dropshipping  via API with DroFxHalmar  dropshipping  via API with DroFx
Halmar dropshipping via API with DroFxolyaivanovalion
 
Log Analysis using OSSEC sasoasasasas.pptx
Log Analysis using OSSEC sasoasasasas.pptxLog Analysis using OSSEC sasoasasasas.pptx
Log Analysis using OSSEC sasoasasasas.pptxJohnnyPlasten
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxolyaivanovalion
 
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Callshivangimorya083
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfMarinCaroMartnezBerg
 
Smarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxSmarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxolyaivanovalion
 
ALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxolyaivanovalion
 
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfAccredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfadriantubila
 
Zuja dropshipping via API with DroFx.pptx
Zuja dropshipping via API with DroFx.pptxZuja dropshipping via API with DroFx.pptx
Zuja dropshipping via API with DroFx.pptxolyaivanovalion
 

Dernier (20)

Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
Best VIP Call Girls Noida Sector 22 Call Me: 8448380779
 
Edukaciniai dropshipping via API with DroFx
Edukaciniai dropshipping via API with DroFxEdukaciniai dropshipping via API with DroFx
Edukaciniai dropshipping via API with DroFx
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
 
CebaBaby dropshipping via API with DroFX.pptx
CebaBaby dropshipping via API with DroFX.pptxCebaBaby dropshipping via API with DroFX.pptx
CebaBaby dropshipping via API with DroFX.pptx
 
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Online
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service OnlineCALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Online
CALL ON ➥8923113531 🔝Call Girls Chinhat Lucknow best sexual service Online
 
Sampling (random) method and Non random.ppt
Sampling (random) method and Non random.pptSampling (random) method and Non random.ppt
Sampling (random) method and Non random.ppt
 
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 night
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 nightCheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 night
Cheap Rate Call girls Sarita Vihar Delhi 9205541914 shot 1500 night
 
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...Vip Model  Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
Vip Model Call Girls (Delhi) Karol Bagh 9711199171✔️Body to body massage wit...
 
Invezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signals
 
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al BarshaAl Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
Al Barsha Escorts $#$ O565212860 $#$ Escort Service In Al Barsha
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptx
 
Halmar dropshipping via API with DroFx
Halmar  dropshipping  via API with DroFxHalmar  dropshipping  via API with DroFx
Halmar dropshipping via API with DroFx
 
Log Analysis using OSSEC sasoasasasas.pptx
Log Analysis using OSSEC sasoasasasas.pptxLog Analysis using OSSEC sasoasasasas.pptx
Log Analysis using OSSEC sasoasasasas.pptx
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFx
 
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
 
FESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdfFESE Capital Markets Fact Sheet 2024 Q1.pdf
FESE Capital Markets Fact Sheet 2024 Q1.pdf
 
Smarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptxSmarteg dropshipping via API with DroFx.pptx
Smarteg dropshipping via API with DroFx.pptx
 
ALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptxALSO dropshipping via API with DroFx.pptx
ALSO dropshipping via API with DroFx.pptx
 
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdfAccredited-Transport-Cooperatives-Jan-2021-Web.pdf
Accredited-Transport-Cooperatives-Jan-2021-Web.pdf
 
Zuja dropshipping via API with DroFx.pptx
Zuja dropshipping via API with DroFx.pptxZuja dropshipping via API with DroFx.pptx
Zuja dropshipping via API with DroFx.pptx
 

Iris Multi-Class Classifier with Azure ML

  • 1. Previously known as Think Big. Move Fast.
  • 2. Iris Multi-Class Classifier with Azure ML A friendly tutorial in 10 Steps V 1.0 – February 2015 Davide Mauri dmauri@solidq.com @mauridb
  • 3. About this tutorial • The objective of this tutorial is to show how to use the original IRIS Dataset with AzureML for multiclass classification • On AzureML the existing dataset is limited to a binary classification… • …but the original one is much more interesting! • We’ll publish the trained model as a web service to be used in your applications • You’ll need a • AzureML Account (free in Preview) • We’ll use a “Supervised Learning” algorithm • Specifically the we’ll be using a Neural Network
  • 4. Useful References • AzureML • https://studio.azureml.net/ • Machine Learning • http://en.wikipedia.org/wiki/Machine_learning • Supervised Learning • http://en.wikipedia.org/wiki/Supervised_learning • Neural Networks • http://en.wikipedia.org/wiki/Artificial_neural_network
  • 5. IRIS Dataset • Available from UC Irvine Machine Learning Repository • http://archive.ics.uci.edu/ml/datasets/Iris • Classification of three Iris species, with four features • Sepal Width & Length, Petal Width & Length • http://en.wikipedia.org/wiki/Iris_flower_data_set • «This is perhaps the best known database to be found in the pattern recognition literature. […] The data set contains 3 classes of 50 instances each, where each class refers to a type of iris plant. One class is linearly separable from the other 2; the latter are NOT linearly separable from each other.» UC Irvine Machine Learning Repository
  • 8. Step 1 – Create Dataset • Create a New Dataset, by uploading the downloaded IRIS Dataset, and name it «Iris UCI Dataset»
  • 10. Step 2 – Create Experiment • Create a new Blank Experiment
  • 11. Step 3 – Add Dataset • Now look for the Iris UCI Dataset in the «Dataset» menu item on the left • Or search it in the top-left search box • Drag and drop it in the design area
  • 12. Step 4 – Add Initialization • Now the model being developed needs to be initialized and trained. • Under the Machine Learning menu look for Initialize Model Classification Multiclass Neural Network and drop it on the design area
  • 13. Step 5 – Split Data • In order to train the model, only a part of the original dataset will be used. The remaining part will be used to evaluate it. • Drop the Split component from the Data Transformation Sample and Split menu and connect the Iris UCI Dataset to the Split Input
  • 14. Step 5 – Split Data • Usually the dataset is split between 70% for training and 30% for evaluating the model. Such configuration can be set in the “Properties” pane.
  • 15. Step 6 – Train Model • Now the Train Model component has to be added in the design area in order to have training working. Drop the component from the Train Train Model element under Machine Learning menu, and connect the item as shown in figure:
  • 16. Step 6 – Train Model • Now select which column will be used as predictors. The Iris Dataset has 5 columns • sepal length in cm • sepal width in cm • petal length in cm • petal width in cm • class • The class column contains the value that we want to be predicted based on the value of the others four columns
  • 17. Step 6 – Train Model • Select the Train Model component that has been placed on the design area before and click on the «Launch columns selector» in the options area and then select the class column
  • 18. Step 6 – Train Model • It’s now possible to run the experiment in order to have the model trained. • Just click on the RUN icon in toolbar to run it • If everything runs correctly a green thick mark will be placed in every component:
  • 19. Step 7 – Save the Trained Model • The model now can be saved as Trained so that it can be used later. • Right click on the Train Model component’s output endpoint • Select «Save as Trained Model» • Name it NN Test
  • 21. Step 8 – Score the Model • Create a new experiment. In the design area add • Iris UCI Dataset • Split component • NN Test from Trained Models • Score Model from Machine Learning Score • Connect • Iris UCI Dataset to Split • Second Split output to second Score Model input • Configure the Split as did before (70/30) • NN Test to first Score Model input
  • 22. Step 9 – Score the Model • Here’s how the resulting experiment should look like:
  • 23. Step 9 – Score the Model • Now you can Run the experiment. This time the trained model will be used to predict the 30% of the data we already know the classification but that wasn’t used in training. • That explain the name of «Supervised Learning». We are teaching the model what we already know so that it can learn how to classify unknown things for use in future • Once the experiment has finished, you can visualize the scored results, by right clicking on the Score Model output and select Visualize
  • 24. Step 9 – Score the Model • In the Visualize window, select the class column and in the «Visualization» pane, in the compare to dropdown, select «Scored Labels» In this case, only 3 out of 45 rows as been wrongly classified! The success rate is above 90%!
  • 26. Step 10 - Publish Web Service • Create a new experiment. In the design area add • Iris UCI Dataset • Project Columns from Data Transformation Manipulation • NN Test from Trained Models • Score Model from Machine Learning Score
  • 27. Step 10 - Publish Web Service • The Project Columns will be used to strip the class column from the data source • Since the model will predict it • It will also define the correct metadata when the model will be published as a Web Service • Connect it with Iris UCI Dataset and with the Score Model • Make sure al but class column are selected in Project Columns properties (use column selector) • Connect the NN Test to the Score Model
  • 28. Step 10 - Publish Web Service • The resulting experiment should look like this
  • 29. Step 10 - Publish Web Service • Run the Experiment • After experiment has finished correctly, add another Project Columns connected to the Score Model • In the column selector select «All Scores» • (At present time) The «All Scores» option is available ONLY after the first run. • This is need for the web service, to strip out all the source columns and keep only the results • Run the experiment again
  • 30. Step 10 - Publish Web Service • The resulting experiment should look like this
  • 31. Step 10 - Publish Web Service • In order to have the Web Service publish, Input and Output data must be defined. • The input will be defined using the metadata of the second Score Model input • Right click on the input and the select «Set as Publish Input»
  • 32. Step 10 - Publish Web Service • The output will be defined using the metadata of the output of the second Project Model component • Right click on the output and the select «Set as Publish Output»
  • 33. Step 10 - Publish Web Service • The resulting experiment should look like this
  • 34. Step 10 - Publish Web Service • Click on the «Publish Web Service» icon • Now the web service can be tested and give sepal and petal data as input, it will return the probability for each class and the most probable class as result • You’ll find the Web Service in the «Web Service» section of AzureML homepage. • Web Service also provides a testing page and examples to use it with • C#, R, Python
  • 36. Conclusion • We created three experiments just for the tutorial purpose but actually only two are needed. • Experiment one and two (Training and Scoring) can be merged together • We only used the Neural Network Classifier but there other Multiclass classifier that could (and should) be used • Test all of them and take the one that give best predictions
  • 37. Previously known as Think Big. Move Fast.

Notes de l'éditeur

  1. Last Changes: 2012-07-30 DM
  2. Last Changes: 2012-07-30 DM