ITCamp 2019 - Silviu Niculita - Supercharge your AI efforts with the use of Automated Machine Learning

@ITCAMPRO #ITCAMP19Community Conference for IT Professionals
Democratizing & Accelerating AI
Through Automated Machine
Learning
Silviu Niculiță
Engineering Manager – Cluj Campus @
Microsoft MVP - AI
silviu@niculita.ro

• Why Machine Learning?
• Why Automated Machine Learning?
• Automated ML Capabilities
• Automated ML Demo
• Conclusions & Call to Action
• Q & A & C
Agenda

WHY MACHINE LEARNING?

ML/AI is becoming a core software capability

Interesting Use Cases
You're probably used to the presence of facial recognition
cameras at airports and other transport hubs, but what about
at concerts?
That's the step Taylor Swift's team took at her May 18th show at
the Rose Bowl, in a bid to identify her stalkers.
According to Rolling Stone, the camera was hidden inside a
display kiosk at the event, and sent images of anyone who
stopped to look at the display to a "command post" in Nashville,
where they were cross-referenced with other photos of the star's
known stalkers.
As the target of numerous death and rape threats, Swift
arguably has a valid motivation for leveraging such technology.
However, it's unclear who has ownership of the photos of her
concertgoers, or how long they will remain on file.

Microsoft's Sports Performance
Platform can, for example, figure out
when a player is at risk of injury,
based on his or her most recent
performance and recovery time.
The difference between a win or a loss
can be decided by an extra five
minutes of wind sprints, levels of
hydration or getting to bed 30
minutes earlier the night before.

The deep learning engine can examine
CT, MRI and other scans and
automatically detect lung, liver, heart
and bone diseases.
New capabilities like lung and breast
cancer, brain trauma, hypertension and
others are "constantly being released,"
the company says.
The results are then passed on to
radiologists, saving them time in
making a diagnosis or requesting
further tests.

“Prison labor” is usually associated
with physical work, but inmates at two
prisons in Finland are doing a new
type of labor: classifying data to train
artificial intelligence algorithms for a
startup.
Though the startup in question, Vainu,
sees the partnership as a kind of
prison reform that teaches valuable
skills, other experts say the claim of
job training is more evidence of hype
around the promises of AI.
Behind The Curtain

WHY AUTOMATED MACHINE LEARNING?

Typical ML Development Process

CURRENT APPROACHES DON’T
SCALE WELL

ITCamp 2019 - Silviu Niculita - Supercharge your AI efforts with the use of Automated Machine Learning

OK….BUT HOW?

ML API Example
How much is this car worth?

Typical ML Creation Example
Mileage
Condition
Car brand
Year of make
Regulations
…
Parameter 1
Parameter 2
Parameter 3
Parameter 4
…
Gradient Boosted
Nearest Neighbors
SVM
Bayesian Regression
LGBM
…
Mileage Gradient Boosted Criterion
Loss
Min Samples Split
Min Samples Leaf
Others Model
Which algorithm? Which parameters?Which features?
Car brand
Year of make

Criterion
Loss
Min Samples Split
Min Samples Leaf
Others
N Neighbors
Weights
Metric
P
Others
Mileage
Condition
Car brand
Year of make
Regulations
…
Gradient Boosted
Nearest Neighbors
SVM
Bayesian Regression
LGBM
…
Nearest Neighbors
Model
Iterate
Gradient BoostedMileage
Car brand
Year of make
Car brand
Year of make
Condition

Iterate

Algorithm Selection Complexity

PUTTING THE AUTO IN AUTO ML

Automated ML Accelerates Model Development
Enter data
Define goals
Apply constraints
OutputInput Intelligently test multiple models in parallel
Optimized model
95%

AUTOMATED ML CAPABILITIES

• ML Scenarios: Classification & Regression;
• Integration: Azure Machine Learning, Azure Notebooks, Jupyter
Notebooks;
• Data Type: Numeric, Text;
• Languages: Python SDK for deployment and hosting for inference;
• Training Compute: Local Machine, Remote Azure DSVM (Linux),
Azure Batch AI;
• Transparency: View run history, model metrics;
• Scale: Faster model training using multiple cores and parallel
experiments;
• Privacy preserving: No need to “see” the data
Automated ML Capabilities

AUTOMATED ML DEMO

CONCLUSIONS & CALL TO ACTION

Target Audience
New to Machine Learning Novice Data Scientists ML Experts
Build model and prefer visual
tool than coding
Quick prototypeLearn machine learning

Manual ML
Dataset
Training
Algorithm 1
Hyperparameter
Values – config 1
Model 1
Hyperparameter
Values – config 2
Model 2
Hyperparameter
Values – config 3
Model 3
Model Training
InfrastructureTraining
Algorithm 2
Hyperparameter
Values – config 4
Model 4

Automated ML
Dataset
Optimization
Metric
Constraints
(Time/Cost)
ML ModelAutomated ML
Accessible & Faster

Competition?

• During training, you only pay for the
cost of the underlying compute;
• When deploying models to AKS, there
is an Azure Machine Learning service
surcharge of $0.040 per core/hr;
• If you deploy to non-AKS VMs, there is
no surcharge;
Pricing

Is Microsoft The Winning Horse?
Gartner 2019 Magic Quadrant for Data Science and Machine
Learning Platforms (as of Nov 2018)
Microsoft remains a Visionary, having maintained a strong
commitment to breadth and ease of open-source technology
integration and excellence in relation to deep learning. Azure
Machine Learning is not an option for the many data science
teams and use cases that require a strictly on-premises
product.

Is AutoML a Silver Bullet?

Start Free
Build, train, and deploy models
with an Azure free account
http://bit.ly/AzureMLS
Documentation
Dig into the technical
documentation
http://bit.ly/AzureMLSDocs
Ask Questions
Send an email to the product team
AskAutomatedML@microsoft.com
Call To Action

• The really nice animated slides in this deck are copied
from a Microsoft AI Platform Team deck;
• Article screenshots are from Engadget, The Verge, The
New York Times and Stanford News;
Credits

Many thanks to our sponsors & partners!
GOLD
SILVER
PARTNERS
PLATINUM
POWERED BY

QUESTIONS & ANSWERS & CANDY
Thank you!

ITCamp 2019 - Silviu Niculita - Supercharge your AI efforts with the use of Automated Machine Learning

More Related Content

What's hot (20)

Similar to ITCamp 2019 - Silviu Niculita - Supercharge your AI efforts with the use of Automated Machine Learning (20)

More from ITCamp (20)

Recently uploaded (20)

ITCamp 2019 - Silviu Niculita - Supercharge your AI efforts with the use of Automated Machine Learning