1.
Azure Machine Learning
BY INDIANDOTNET A MICROSOFT USER COMMUNITY GROUP

2.
Agenda
 What is Machine Learning ?
 How does Machine Learning Help us ?
 Different Type of Machine Learning
 Different techniques or Algorithm to Solve Problem ?
 Azure Machine Learning
 Hands on Azure ML Studio

3.
Machine Learning
 “A breakthrough in machine learning would be worth ten
Microsoft” - Bill Gates, Chairman, Microsoft
 “Machine learning is the next Internet” - Tony Tether, former director ,
DARPA
 “Machine learning is the hot new thing” - John Hennessy, President,
Stanford
 “Web rankings today are mostly a matter of machine learning”
Prabhakar Raghavan, former Dir. Research, Yahoo
 “Machine learning is going to result in a real revolution” –
Greg Papadopoulos, former CTO, Sun

4.
What is Machine Learning?
Machine learning is a way to understand the data pattern , recognize it and
predict accordingly for future. It helps in
Data Mining, Language Processing, Image recognition and many other Artificial
Intelligence
And below is from Wikipedia
Tom M. Mitchell provided a widely quoted, more formal definition: "A computer
program is said to learn from experience E with respect to some class of tasks T
and performance measure P if its performance at tasks in T, as measured by P,
improves with experience E."

5.
How does Machine Learning Help us ?
 The United States Postal Service
processed over 150 billion pieces of mail
in 2013—far too much for efficient
human sorting.
 But as recently as 1997, only 10% of
hand-addressed mail was successfully
sorted automatically.

6.
Aap likhe khuda vache – Different hand
writing
Biggest challenge as handwriting
may vary person to person from
worst to best or best to worst.

7.
How does Machine Learning Help ?
Credit card Fraud
detection Face detection
Search
recommendation

8.
Weather
forecasting

9.
Machine Learning
Although, Machine learning Is more than this. Here, we are showing
some more example where machine learning can help
 Determine SPAM emails
 Provide customer like to switch to competitor
 Free text when typing
 Xbox gaming
And many more examples.

10.
“
”
Types of Machine Learning

11.
Supervised Learning
 The Supervised learning means the value you want to predict is
already exist in training data. Means the data already exist in the
computer so data is labeled. The accuracy is high in such case.
 Used when you want to find unknown answers and have data with
known answers

12.
Car

13.
Hey I am a car ?

14.
Unsupervised Machine Learning
 It is just opposite to Supervised Machine Learning. In this
the predictive data not present in training data.
 It is Used when you want to find unknown answers – mostly
groupings – directly from data
 No simple way to evaluate accuracy

15.
“
”
Algorithm Class
Different techniques for solving the problem

16.
classification
Binary classification (Two
Class )
– yes/no, 1/0,
male/female

17.
Regression
 Predict a real value – temperature, stock value...

18.
Anomaly detection

19.
Clustering
 Partition items into group – twitter posts

20.
“
”
Azure Machine Learning

21.
Azure Machine Learning
 Newest Azure service which reduces complexity of Machine learning
 process and brings ML to broader audience
 Possibility to develop and put into production ML models without
 writing line of code.
 Availability of many top of the class machine learning algorithms
 (internally used at other Microsoft products)
 Easy ML model deployment and usage using restful API
 Easy collaboration on Azure Machine Learning projects
 Support for open source framework R

22.
Work Flow

23.
“
”
First Look Of Azure Machine
Learning Studio
https://studio.azureml.net

24.
Thank you
Rajat Jaiswal (Microsoft MVP)
http://nerdtechies.com
http://indiandotnet.wordpress.com
http://facebook.com/indiandotnet