Вы находитесь на странице: 1из 21

Machine Learning

Integrating Power BI and Azure Machine Learning


Agenda
Introductions
Terms and definitions
Why is this technology relevant
What is Azure Machine Learning
What is Power BI
Demo
Use Cases
Q&A
Who am I A bit about me….
• Working exclusively on BI projects for over 12 years,
over 15 in IT

• Worked in many different industries both public and


private

• Co-organizer of Power BI User group in Victoria

• Partner Seller for Microsoft

• Presented at Microsoft BUILD 2017

Anthony Bulk • Presented at Power BI World Tour 2017

Principal • Solution architect on an award winning AI solution


• Impact award 2017
• Gold on CDN IoT Solution of the Year 2017
• Gold on CDN Disruptor of the Year 2017
Sierra Systems
• Member of the following MS Advisory Groups
• Integration Services (SSIS) Advisors
• Power BI Advisors
• Time Series Insights Preview
• Cortana Intelligence Advisor
• Big Data Advisors
• Data Factory Advisors
• Stream Analytics Advisors
• Machine Learning Advisors
Terms and definitions
A.I. Artificial or Augmented Intelligence Deep Learning
Deep learning is focused on training a neural
Artificial intelligence (AI) is the intelligence
network that learns as it is exposed to more
exhibited by machines or software. It includes
and more data. The network adapts based on
Machine Learning and Deep learning. features in the raw data as opposed to
machine learning in which features must be
engineered that the algorithm can use.

Machine Learning
Machine learning focuses on the Cognitive Computing
development of computer programs The definition is still evolving
that can change when exposed to new however the general consensus is
data. Typically the methods and that it is hardware and/or software
algorithms used in this area are task- that can mimic the functions of the
specific. human brain.
Why is this technology relevant?

“It’s hard to overstate how big of an impact AI is going to have on society over the
next 20 years” — Jeff Bezos, CEO, Amazon

“Within a few years, every major decision – personal or business – will be made with
the help of AI and cognitive technologies” — Ginni Rometty, CEO, IBM

“The business plans of the next 10,000 start-ups are easy to forecast: Take X and
add AI. This is a big deal, and now it’s here” — Kevin Kelly, Founding Executive Editor,
Wired magazine

The technology exists today. It is cheap, fast and reliable. Organizations that leverage
these technologies will be able to obtain or sustain a competitive advantage. Those
that don't will be left behind.
Difficulty vs Business Value - Before
How can we make it happen?

Prescriptive Analytics

What will happen?

Predictive Analytics

Why did it happen?


Barrier to entry
Diagnostic Analytics
DIFFICULTY

What happened?

Descriptive Analytics

Traditional BI Advanced Analytics

BUSINESS VALUE
Difficulty vs Business Value - Now

The line between traditional BI


and Advanced Analytics is Demand and
getting blurred technology
advances is
lowering the bar

How can we make it happen?


DIFFICULTY

What will happen? Prescriptive Analytics

Why did it happen? Predictive Analytics

What happened? Diagnostic Analytics

Descriptive Analytics Barrier to entry

The gap is
shrinking,
are you
ready?

Traditional BI Advanced Analytics

BUSINESS VALUE
The decision support continuum

Operational Reporting Business Intelligence Artificial Intelligence


The decision support continuum

Streaming Predictive Analytics


Analytics
Operational Reporting Business Intelligence Prescriptive
Artificial Intelligence
Real Time Analytics Analytics
The decision support continuum

Operational Reporting Business Intelligence Artificial Intelligence


What is Azure Machine Learning and Power BI?
Cortana Intelligence Suite
What is Azure Machine Learning?

• Azure Machine Learning is a cloud predictive analytics service that makes it


possible to quickly create and deploy predictive models as analytics solutions. You
can work from a ready-to-use library of algorithms, use them to create models on
an internet-connected PC, and deploy your predictive solution quickly.

• Microsoft Azure Machine Learning Studio is a collaborative, drag-and-drop tool


you can use to build, test, and deploy predictive analytics solutions on your
data. Machine Learning Studio publishes models as web services that can easily
be consumed by custom apps or BI tools such as Excel.
What is Power BI?
Power BI and Azure Machine Learning - Overview

Azure Machine
Learning

Cloud
On Premise

Power BI Desktop
SQL Server
Demo
Power BI and Azure Machine Learning - Workflow
Step 5 1
2
3
4
The web
Save
Create theonPower
aanR
service
RODBC
data
script
BI
frame
connection
will
that
report
return
of
willand
be
atoadd
yourRSQL
used
JSON
prediction
the as
result
script
a Server
data
results
as
setasource
todata
database.
data
thesource
in
Rback
script.
Power
This
to
first.
BI.
ODBC
In
This
SQL
Then
theServer
JSON
add
connection
R script
the
result
using
table
callset
the
will
theof
will
same
prediction
be
web
need
used
ODBC
service
to
in
R toloaded
and
be
connection
results
send
extract
in SQL
itinto
the
used
the
Server.
adata
data
data
to extract
which
out
Typically
frame
of the
SQL
andI
Server.
model
formatted
data
like to include
will
Thissopredict
data
that
bothwill
itvalues
the
can
thendata
befor.
be
saved
that
passed
to
wasa table
sent upupin
toto
SQL
thetheServer.
Azure
Azure Machine
Machine
Learning model
Web Service as welltoasmakethe
predictionsresults
prediction on. that the Azure
Machine Learning model
generated in the Power BI Report.
Power BI and Azure Machine Learning - Overview

Enterprise Solution

Execution of the R script can


be shifted out of Power BI
and into SQL Server 2016.

This centralizes and secures


the integration code and
moves it into the domain of
IT to manage.
Uses Cases

Predictive Maintenance
Anticipate equipment failures before they happen and prevent then from happening.

Resource Planning
Allocate the correct staffing levels at the right place at the right time by predicting
resource needs

Employee retention
Identify employees that are a flight risk and intervene before they are lost

Supply Chain Management


Minimize disruptions to the supply chain by predicting which products may run out
or flagging suppliers that are at risk of late delivery before it happens.
When to use what? Microsoft ML & AI
products

Build your own or consume Build your own Consume


pre-trained models?

Azure Machine Learning Cognitive services, bots

Which Experience do you Code first Visual tooling


want?

On-premise Cloud Cloud


Deployment target AML Studio
ML Server AML Services (Preview)

On-premise SQL SQL Spark Hadoop Azure Batch DSVM Azure Container
What engine(s) do you want
Hadoop Server Server Service
to use?
Questions
Thank You
Email: AnthonyBulk@sierrasystems.com
Twitter: @JAnthonyBulk
LinkedIn: https://www.linkedin.com/in/janthonybulk/

Вам также может понравиться