SlideShare a Scribd company logo
1 of 55
Deep learning
made doubly easy with
reusable deep features
Carlos Guestrin
Dato, CEO
University of Washington, Amazon Prof. of ML
Successful apps
in 2015 must be
intelligent
Machine
learning
key to next-gen apps
ā€¢ Recommenders
ā€¢ Fraud detection
ā€¢ Ad targeting
ā€¢ Financial models
ā€¢ Personalized medicine
ā€¢ Churn prediction
ā€¢ Smart UX
(video & text)
ā€¢ Personal assistants
ā€¢ IoT
ā€¢ Socials nets
ā€¢ ā€¦Last decade:
Data management
Now:
Intelligent apps
?
Last 5 years:
Traditional analytics
The ML pipeline circa 2013
DATA
ML
Algorithm
My curve is
better than
your curve
Write a
paper
2015: Production ML pipeline
DATA
YourWebServiceor
IntelligentApp
ML
Algorithm
Data
cleaning
&
feature
eng
Offline
eval &
Parameter
search
Deploy
model
Data engineering Data intelligence Deployment
Using deep learning
Goal: Platform to help implement, manage, optimize entire pipeline
Todayā€™s talk
Features in
ML
Neural
networks
Deep
learning for
computer
vision
Deep
learning
made easy
with deep
features
Applications
to text data
Deployment
in
production
Features are key to machine learning
7
Simple example: Spam filtering
ā€¢ A user walks into an emailā€¦
- Will she thinks its spam?
ā€¢ Whatā€™s the probability email is spam?
Text of email
User info
Source info
Input: x
MODEL
Yes!
No
Output:
Probability of y
8
Feature engineering:
the painful black art of transforming raw inputs
into useful inputs for ML algorithm
ā€¢ E.g., important words, stemming text, complex
transformation of inputs,ā€¦
MODEL
Yes!
No
Output:
Probability of y
Feature
extraction
Features: Ī¦(x)
Text of email
User info
Source info
Input: x
Neural networks
ļƒ§ļƒØ
Learning *very* non-linear features
10
Linear classifiers
ā€¢ Most common classifier
- Logistic regression
- SVMs
- ā€¦
ā€¢ Decision correspond to
hyperplane:
- Line in high dimensional
space
w0 + w1 x1 + w2 x2 > 0 w0 + w1 x1 + w2 x2 < 0
11
Graph representation of classifier:
useful for defining neural networks
x
1
x
2
x
d
y
ā€¦
1
w2 w0 + w1 x1 + w2 x2 + ā€¦ + wd xd
> 0, output 1
< 0, output 0
Input Output
12
What can a linear classifier represent
x1 OR x2 x1 AND x2
x
1
x
2
1
y
-0.5
1
1
x
1
x
2
1
y
-1.5
1
1
13
What canā€™t a simple linear classifier represent?
XOR
the counterexample
to everything
Need non-linear features
Solving the XOR problem: Adding a layer
XOR = x1 AND NOT x2 OR NOT x1 AND x2
z
1
-0.5
1
-1
z1 z2
z
2
-0.5
-1
1
x
1
x
2
1
y
1 -0.5
1
1
Thresholded to 0 or 1
15
A neural network
ā€¢ Layers and layers and layers of linear models and non-linear
transformation
ā€¢ Around for about 50 years
- Fell in ā€œdisfavorā€ in 90s
ā€¢ In last few years, big resurgence
- Impressive accuracy on a several benchmark problems
- Powered by huge datasets, GPUs, & modeling/learning alg
improvements
x
1
x
2
1
z
1
z
2
1
y
Applications to computer vision
(or the deep devil is in the deep details)
17
Image features
ā€¢ Features = local detectors
- Combined to make prediction
- (in reality, features are more low-level)
Face!
Eye
Eye
Nose
Mouth
18
Many hand create features existā€¦
Computer$vision$features$
SIFT$ Spin$image$
HoG$ RIFT$
Textons$ GLOH$
Slide$Credit:$Honglak$Lee$
19
Standard image classification approach
Input
Computer$vision$features$
SIFT$ Spin$image$
HoG$ RIFT$
Textons$ GLOH$
Slide$Credit:$Honglak$Lee$
Extract features Use simple classifier
e.g., logistic regression, SVMs
Car?
20
Many hand create features existā€¦
Computer$vision$features$
SIFT$ Spin$image$
HoG$ RIFT$
Textons$ GLOH$
Slide$Credit:$Honglak$Lee$
ā€¦ but very painful to design
21
Use neural network to learn features
Each layer learns features, at different levels of abstraction
Y LeCun
MA Ranzato
Deep Learning = Learning Hierarchical Representations
It's deep if it has more than one stage of non-linear feature
transformation
Trainable
Classifier
Low-Level
Feature
Mid-Level
Feature
High-Level
Feature
Feature visualization of convolutional net trained on ImageNet from [ Zeiler & Fergus 2013]
22
Many tricks needed to work wellā€¦
ā€¢ Different types of layers, connections,ā€¦ needed for high accuracy
Krizhevsky et al.
ā€˜12
Sample performance results
Sample results
ā€¢ Traffic sign recognition
(GTSRB)
- 99.2% accuracy
ā€¢ House number recognition
(Google)
- 94.3% accuracy
30
Krizhevsky et al. ā€™12:
60M parameters, won 2012 ImageNet competition
31
32
ImageNet 2012 competition: 1.5M images, 1000 categories
32
33
Ā©Carlos Guestrin 2005-2014
33
MA
TEST
IMAGE RETRIEVED IMAGES
34
Application to scene parsing
Ā©Carlos Guestrin 2005-2014
Y LeCun
MA Ranzato
Semantic Labeling:
Labeling every pixel with the object it belongs to
[ Farabet et al. ICML 2012, PAMI 2013]
Would help identify obstacles, targets, landing sites, dangerous areas
Would help line up depth map with edge maps
Challenges of deep learning
Deep learning score card
Pros
ā€¢ Enables learning of features rather
than hand tuning
ā€¢ Impressive performance gains on
- Computer vision
- Speech recognition
- Some text analysis
ā€¢ Potential for much more impact
Cons
Deep learning workflow
Lots of
labeled data
Training set
Validation set
80%
20%
Learn deep
neural net
model
Validate
Deep learning score card
Pros
ā€¢ Enables learning of features rather
than hand tuning
ā€¢ Impressive performance gains on
- Computer vision
- Speech recognition
- Some text analysis
ā€¢ Potential for much more impact
Cons
ā€¢ Computationally really expensive
ā€¢ Requires a lot of data for high
accuracy
ā€¢ Extremely hard to tune
- Choice of architecture
- Parameter types
- Hyperparameters
- Learning algorithm
- ā€¦
ā€¢ Computational + so many choices =
incredibly hard to tune
Deep features:
Deep learning
+
Transfer learning
40
Change image classification approach?
Input
Computer$vision$features$
SIFT$ Spin$image$
HoG$ RIFT$
Textons$ GLOH$
Slide$Credit:$Honglak$Lee$
Extract features Use simple classifier
e.g., logistic regression, SVMs
Car?
Can we learn features
from data,
even when
we donā€™t have
data or time?
41
Transfer learning:
Use data from one domain to help learn on another
Lots of data:
Learn
neural net
Great
accuracy on
cat v. dogvs.
Some data: Neural net as
feature extractor
+
Simple classifier
Great accuracy on
101 categories
Old idea, explored for deep learning by Donahue et al. ā€™14
42
Whatā€™s learned in a neural net
Neural net trained for Task 1: cat vs. dog
Very specific to Task 1
Should be ignored for other tasks
More generic
Can be used as feature extractor
vs.
43
Transfer learning in more detailā€¦
Neural net trained for Task 1: cat vs. dog
Very specific to Task 1
Should be ignored for other tasks
More generic
Can be used as feature extractor
Keep weights fixed!
For Task 2, predicting 101 categories, learn only end part
Use simple classifier
e.g., logistic regression, SVMs
Class
?
44
Careful where you cutā€¦
Last few layers tend to be too specific
Y LeCun
MA Ranzato
Deep Learning = Learning Hierarchical Representations
It's deep if it has more than one stage of non-linear feature
transformation
Trainable
Classifier
Low-Level
Feature
Mid-Level
Feature
High-Level
Feature
Feature visualization of convolutional net trained on ImageNet from [ Zeiler & Fergus 2013]
Too specific for
car detectionUse these!
Transfer learning with deep features
Training set
Validation set
80%
20%
Learn
simple
model
Some
labeled data
Extract
features with
neural net
trained on
different task
Validate
Deploy in
production
How general are deep features?
Applications to text data
Simple text classification with bag of words
aardvark 0
about 2
all 2
Africa 1
apple 0
anxious 0
...
gas 1
...
oil 1
ā€¦
Zaire 0
Use simple classifier
e.g., logistic regression, SVMs
Class
?
One ā€œfeatureā€ per word
Word2Vec: Neural network for finding high
dimensional representation per word Mikolov et al. ā€˜13
Skip-gram Model: From a word, predict nearby words in sentence
Awesome learning talk at
Strata
deep
300 dim
representation
300 dim
representation
300 dim
representation
300 dim
representation
300 dim
representation
300 dim
representation
Neural net
Viewed as deep
features
50
Related words placed nearby high dim space
Projecting 300 dim space into 2 dim with PCA (Mikolov et al. ā€™13)
Classifier:
e.g., logistic regression, SVMs with
300 x number_of_words parameters
Class
?
Embed each
word into
300 dim
space
Text classification with word embeddings
aardvark 0
about 2
all 2
Africa 1
apple 0
anxious 0
...
gas 1
...
oil 1
ā€¦
Zaire 0
Practical example
Blog corpus
Haha
Yea
Hahaha
Hahah
Lisxc
Umm
Hehe
laughingoutloud
LOL
Closest words
in 300 dim
Predicts gender of author with 79% accuracy
Deploying ML in production
55
DATA
ML
Algorithm
Deployment?
ā€¢ Write spec, other team
implements in ā€˜productionā€™ language
o 6-12 months
o Stale/irrelevant model/approach
o 2 teams maintaining 2 systems
Custom
Model
Data Engineers, Data
Architects, DevOps,
App Developers
App
A
P
I
Data Scientist
ML deployment requirements
56
Easy to
integrate
Rest API
Scalable
Fault tolerant
Flexible
Any model,
any Python
App
A
P
I
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
LB
GLC
Model
GLC
Model
GLC
Model
Dato
Models
Dato
Models
Dato
Models
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
LB
GLC
Model
GLC
Model
GLC
Model
Python
Python
Python
57
Do-It-Yourself
ā€¢ Web Service layer:
- Tornado, Flask, Keen, Django, ā€¦
ā€¢ Caching layer:
- Redis, Cassandra, Memcached,
DynamoDb, MySQL, ā€¦
ā€¢ Logs:
- Logback, LogStash, Splunk, Loggly, ā€¦
ā€¢ Metrics:
- AWS CloudWatch, Mixpanel, Librato, ā€¦
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
A
P
I
C
A
C
H
E
LB
GLC
Model
GLC
Model
GLC
Model
Python
Python
Python
App
58
ā€¦ or use Dato Predictive Services
YourWebServiceor
IntelligentApp
ML Model
Dato
Predictive Services
CachingLayer
Predictive
ObjectServer
Serves predictions in a robust, scalable, incremental fashion
Better
ML Model
Serve any model: GraphLab Create, scikit-learn, Python, ā€¦
ā€¢ Out-of-core computation
ā€¢ Tools for feature engineering
ā€¢ Rich data type support
ā€¢ Models built for scale
ā€¢ App-oriented toolkits
ā€¢ Advanced ML & Extensible
ā€¢ Deploy models as low-latency REST services
ā€¢ Same code for distributed computation
ā€¢ Elastically scale up or out with one command
ā€¢ Job monitoring & model management
ā€¢ Deploy existing Python code & models
ā€¢ Run on AWS EC2 or Hadoop Yarn
SGraph
Create Engine
SFrameCanvas
Machine Learning Toolkits SDK
GraphLab Create Dato Distributed Dato Predictive Services
Predictive Engine
REST Client Direct
Model Mgmt
Distributed Engine
DirectJob Client
Job Mgmt
Dato Platform
Summary
Deep learning made easy with deep features
Deep learning: exciting ML development
Slow, lots of tuning,
needs lots of data
Deep features: reuse deep models for new domains
Needs less data Faster training times Much simpler tuning
Can still achieve excellent performance

More Related Content

What's hot

NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digitsNVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
NVIDIA Taiwan
Ā 

What's hot (20)

AI powered emotion recognition: From Inception to Production - Global AI Conf...
AI powered emotion recognition: From Inception to Production - Global AI Conf...AI powered emotion recognition: From Inception to Production - Global AI Conf...
AI powered emotion recognition: From Inception to Production - Global AI Conf...
Ā 
MXNet Workshop
MXNet WorkshopMXNet Workshop
MXNet Workshop
Ā 
AI powered emotion recognition: From Inception to Production - Global AI Conf...
AI powered emotion recognition: From Inception to Production - Global AI Conf...AI powered emotion recognition: From Inception to Production - Global AI Conf...
AI powered emotion recognition: From Inception to Production - Global AI Conf...
Ā 
專锌報告
專锌報告專锌報告
專锌報告
Ā 
The Search for a New Visual Search Beyond Language - StampedeCon AI Summit 2017
The Search for a New Visual Search Beyond Language - StampedeCon AI Summit 2017The Search for a New Visual Search Beyond Language - StampedeCon AI Summit 2017
The Search for a New Visual Search Beyond Language - StampedeCon AI Summit 2017
Ā 
Introduction to Deep Learning and neon at Galvanize
Introduction to Deep Learning and neon at GalvanizeIntroduction to Deep Learning and neon at Galvanize
Introduction to Deep Learning and neon at Galvanize
Ā 
The deep learning tour - Q1 2017
The deep learning tour - Q1 2017 The deep learning tour - Q1 2017
The deep learning tour - Q1 2017
Ā 
Generative Models for General Audiences
Generative Models for General AudiencesGenerative Models for General Audiences
Generative Models for General Audiences
Ā 
Using Deep Learning to do Real-Time Scoring in Practical Applications
Using Deep Learning to do Real-Time Scoring in Practical ApplicationsUsing Deep Learning to do Real-Time Scoring in Practical Applications
Using Deep Learning to do Real-Time Scoring in Practical Applications
Ā 
An introduction to Machine Learning (and a little bit of Deep Learning)
An introduction to Machine Learning (and a little bit of Deep Learning)An introduction to Machine Learning (and a little bit of Deep Learning)
An introduction to Machine Learning (and a little bit of Deep Learning)
Ā 
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr..."Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
"Trade-offs in Implementing Deep Neural Networks on FPGAs," a Presentation fr...
Ā 
COCOA: Communication-Efficient Coordinate Ascent
COCOA: Communication-Efficient Coordinate AscentCOCOA: Communication-Efficient Coordinate Ascent
COCOA: Communication-Efficient Coordinate Ascent
Ā 
Squeezing Deep Learning Into Mobile Phones
Squeezing Deep Learning Into Mobile PhonesSqueezing Deep Learning Into Mobile Phones
Squeezing Deep Learning Into Mobile Phones
Ā 
NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digitsNVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
NVIDIA ę·±åŗ¦å­øēæ’ę•™č‚²ę©Ÿę§‹ (DLI): Medical image segmentation using digits
Ā 
Android and Deep Learning
Android and Deep LearningAndroid and Deep Learning
Android and Deep Learning
Ā 
Apache MXNet ODSC West 2018
Apache MXNet ODSC West 2018Apache MXNet ODSC West 2018
Apache MXNet ODSC West 2018
Ā 
Deep Learning Primer: A First-Principles Approach
Deep Learning Primer: A First-Principles ApproachDeep Learning Primer: A First-Principles Approach
Deep Learning Primer: A First-Principles Approach
Ā 
Keynote: Machine Learning for Design Automation at DAC 2018
Keynote:  Machine Learning for Design Automation at DAC 2018Keynote:  Machine Learning for Design Automation at DAC 2018
Keynote: Machine Learning for Design Automation at DAC 2018
Ā 
Evolving a Medical Image Similarity Search
Evolving a Medical Image Similarity SearchEvolving a Medical Image Similarity Search
Evolving a Medical Image Similarity Search
Ā 
Introduction to Deep Learning with Will Constable
Introduction to Deep Learning with Will ConstableIntroduction to Deep Learning with Will Constable
Introduction to Deep Learning with Will Constable
Ā 

Similar to Strata London - Deep Learning 05-2015

Deep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep FeaturesDeep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep Features
Turi, Inc.
Ā 
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr TeterwakLearn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
PyData
Ā 
Lessons Learned from Building Machine Learning Software at Netflix
Lessons Learned from Building Machine Learning Software at NetflixLessons Learned from Building Machine Learning Software at Netflix
Lessons Learned from Building Machine Learning Software at Netflix
Justin Basilico
Ā 
OSCON 2014: Data Workflows for Machine Learning
OSCON 2014: Data Workflows for Machine LearningOSCON 2014: Data Workflows for Machine Learning
OSCON 2014: Data Workflows for Machine Learning
Paco Nathan
Ā 
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr..."Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
Edge AI and Vision Alliance
Ā 

Similar to Strata London - Deep Learning 05-2015 (20)

Deep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep FeaturesDeep Learning Made Easy with Deep Features
Deep Learning Made Easy with Deep Features
Ā 
How to use transfer learning to bootstrap image classification and question a...
How to use transfer learning to bootstrap image classification and question a...How to use transfer learning to bootstrap image classification and question a...
How to use transfer learning to bootstrap image classification and question a...
Ā 
OReilly AI Transfer Learning
OReilly AI Transfer LearningOReilly AI Transfer Learning
OReilly AI Transfer Learning
Ā 
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr TeterwakLearn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Learn to Build an App to Find Similar Images using Deep Learning- Piotr Teterwak
Ā 
Deep Learning with CNTK
Deep Learning with CNTKDeep Learning with CNTK
Deep Learning with CNTK
Ā 
The Frontier of Deep Learning in 2020 and Beyond
The Frontier of Deep Learning in 2020 and BeyondThe Frontier of Deep Learning in 2020 and Beyond
The Frontier of Deep Learning in 2020 and Beyond
Ā 
B4UConference_machine learning_deeplearning
B4UConference_machine learning_deeplearningB4UConference_machine learning_deeplearning
B4UConference_machine learning_deeplearning
Ā 
Build, Train, and Deploy ML Models at Scale
Build, Train, and Deploy ML Models at ScaleBuild, Train, and Deploy ML Models at Scale
Build, Train, and Deploy ML Models at Scale
Ā 
The Data Science Process - Do we need it and how to apply?
The Data Science Process - Do we need it and how to apply?The Data Science Process - Do we need it and how to apply?
The Data Science Process - Do we need it and how to apply?
Ā 
C19013010 the tutorial to build shared ai services session 1
C19013010  the tutorial to build shared ai services session 1C19013010  the tutorial to build shared ai services session 1
C19013010 the tutorial to build shared ai services session 1
Ā 
Lessons Learned from Building Machine Learning Software at Netflix
Lessons Learned from Building Machine Learning Software at NetflixLessons Learned from Building Machine Learning Software at Netflix
Lessons Learned from Building Machine Learning Software at Netflix
Ā 
Introduction to ML.NET
Introduction to ML.NETIntroduction to ML.NET
Introduction to ML.NET
Ā 
OSCON 2014: Data Workflows for Machine Learning
OSCON 2014: Data Workflows for Machine LearningOSCON 2014: Data Workflows for Machine Learning
OSCON 2014: Data Workflows for Machine Learning
Ā 
Amazon SageMaker 內å»ŗę©Ÿå™Øå­øēæ’ę¼”ē®—ę³• (Level 400)
Amazon SageMaker 內å»ŗę©Ÿå™Øå­øēæ’ę¼”ē®—ę³• (Level 400)Amazon SageMaker 內å»ŗę©Ÿå™Øå­øēæ’ę¼”ē®—ę³• (Level 400)
Amazon SageMaker 內å»ŗę©Ÿå™Øå­øēæ’ę¼”ē®—ę³• (Level 400)
Ā 
Data Workflows for Machine Learning - Seattle DAML
Data Workflows for Machine Learning - Seattle DAMLData Workflows for Machine Learning - Seattle DAML
Data Workflows for Machine Learning - Seattle DAML
Ā 
Prepare your data for machine learning
Prepare your data for machine learningPrepare your data for machine learning
Prepare your data for machine learning
Ā 
Machine learning at scale - Webinar By zekeLabs
Machine learning at scale - Webinar By zekeLabsMachine learning at scale - Webinar By zekeLabs
Machine learning at scale - Webinar By zekeLabs
Ā 
Big Data Analytics (ML, DL, AI) hands-on
Big Data Analytics (ML, DL, AI) hands-onBig Data Analytics (ML, DL, AI) hands-on
Big Data Analytics (ML, DL, AI) hands-on
Ā 
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr..."Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
"Enabling Ubiquitous Visual Intelligence Through Deep Learning," a Keynote Pr...
Ā 
Designing Artificial Intelligence
Designing Artificial IntelligenceDesigning Artificial Intelligence
Designing Artificial Intelligence
Ā 

More from Turi, Inc.

Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge DatasetsScaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Turi, Inc.
Ā 
Machine Learning in Production with Dato Predictive Services
Machine Learning in Production with Dato Predictive ServicesMachine Learning in Production with Dato Predictive Services
Machine Learning in Production with Dato Predictive Services
Turi, Inc.
Ā 

More from Turi, Inc. (20)

Webinar - Analyzing Video
Webinar - Analyzing VideoWebinar - Analyzing Video
Webinar - Analyzing Video
Ā 
Webinar - Patient Readmission Risk
Webinar - Patient Readmission RiskWebinar - Patient Readmission Risk
Webinar - Patient Readmission Risk
Ā 
Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)Webinar - Know Your Customer - Arya (20160526)
Webinar - Know Your Customer - Arya (20160526)
Ā 
Webinar - Product Matching - Palombo (20160428)
Webinar - Product Matching - Palombo (20160428)Webinar - Product Matching - Palombo (20160428)
Webinar - Product Matching - Palombo (20160428)
Ā 
Webinar - Pattern Mining Log Data - Vega (20160426)
Webinar - Pattern Mining Log Data - Vega (20160426)Webinar - Pattern Mining Log Data - Vega (20160426)
Webinar - Pattern Mining Log Data - Vega (20160426)
Ā 
Webinar - Fraud Detection - Palombo (20160428)
Webinar - Fraud Detection - Palombo (20160428)Webinar - Fraud Detection - Palombo (20160428)
Webinar - Fraud Detection - Palombo (20160428)
Ā 
Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge DatasetsScaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Scaling Up Machine Learning: How to Benchmark GraphLab Create on Huge Datasets
Ā 
Pattern Mining: Extracting Value from Log Data
Pattern Mining: Extracting Value from Log DataPattern Mining: Extracting Value from Log Data
Pattern Mining: Extracting Value from Log Data
Ā 
Intelligent Applications with Machine Learning Toolkits
Intelligent Applications with Machine Learning ToolkitsIntelligent Applications with Machine Learning Toolkits
Intelligent Applications with Machine Learning Toolkits
Ā 
Text Analysis with Machine Learning
Text Analysis with Machine LearningText Analysis with Machine Learning
Text Analysis with Machine Learning
Ā 
Machine Learning with GraphLab Create
Machine Learning with GraphLab CreateMachine Learning with GraphLab Create
Machine Learning with GraphLab Create
Ā 
Machine Learning in Production with Dato Predictive Services
Machine Learning in Production with Dato Predictive ServicesMachine Learning in Production with Dato Predictive Services
Machine Learning in Production with Dato Predictive Services
Ā 
Machine Learning in 2016: Live Q&A with Carlos Guestrin
Machine Learning in 2016: Live Q&A with Carlos GuestrinMachine Learning in 2016: Live Q&A with Carlos Guestrin
Machine Learning in 2016: Live Q&A with Carlos Guestrin
Ā 
Scalable data structures for data science
Scalable data structures for data scienceScalable data structures for data science
Scalable data structures for data science
Ā 
Introduction to Deep Learning for Image Analysis at Strata NYC, Sep 2015
Introduction to Deep Learning for Image Analysis at Strata NYC, Sep 2015Introduction to Deep Learning for Image Analysis at Strata NYC, Sep 2015
Introduction to Deep Learning for Image Analysis at Strata NYC, Sep 2015
Ā 
Introduction to Recommender Systems
Introduction to Recommender SystemsIntroduction to Recommender Systems
Introduction to Recommender Systems
Ā 
Machine learning in production
Machine learning in productionMachine learning in production
Machine learning in production
Ā 
Overview of Machine Learning and Feature Engineering
Overview of Machine Learning and Feature EngineeringOverview of Machine Learning and Feature Engineering
Overview of Machine Learning and Feature Engineering
Ā 
SFrame
SFrameSFrame
SFrame
Ā 
Building Personalized Data Products with Dato
Building Personalized Data Products with DatoBuilding Personalized Data Products with Dato
Building Personalized Data Products with Dato
Ā 

Recently uploaded

Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
amitlee9823
Ā 
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service BangaloreCall Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
amitlee9823
Ā 
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
AroojKhan71
Ā 
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
amitlee9823
Ā 
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICECHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
9953056974 Low Rate Call Girls In Saket, Delhi NCR
Ā 
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
MarinCaroMartnezBerg
Ā 
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
amitlee9823
Ā 
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
amitlee9823
Ā 
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
amitlee9823
Ā 

Recently uploaded (20)

Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
Chintamani Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore ...
Ā 
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service BangaloreCall Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
Call Girls Begur Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service Bangalore
Ā 
(NEHA) Call Girls Katra Call Now 8617697112 Katra Escorts 24x7
(NEHA) Call Girls Katra Call Now 8617697112 Katra Escorts 24x7(NEHA) Call Girls Katra Call Now 8617697112 Katra Escorts 24x7
(NEHA) Call Girls Katra Call Now 8617697112 Katra Escorts 24x7
Ā 
VidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptxVidaXL dropshipping via API with DroFx.pptx
VidaXL dropshipping via API with DroFx.pptx
Ā 
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
Ā 
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Bommasandra Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Ā 
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICECHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
CHEAP Call Girls in Saket (-DELHI )šŸ” 9953056974šŸ”(=)/CALL GIRLS SERVICE
Ā 
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
Ā 
Anomaly detection and data imputation within time series
Anomaly detection and data imputation within time seriesAnomaly detection and data imputation within time series
Anomaly detection and data imputation within time series
Ā 
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
Junnasandra Call Girls: šŸ“ 7737669865 šŸ“ High Profile Model Escorts | Bangalore...
Ā 
Halmar dropshipping via API with DroFx
Halmar  dropshipping  via API with DroFxHalmar  dropshipping  via API with DroFx
Halmar dropshipping via API with DroFx
Ā 
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
Ā 
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
Ā 
Week-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interactionWeek-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interaction
Ā 
ELKO dropshipping via API with DroFx.pptx
ELKO dropshipping via API with DroFx.pptxELKO dropshipping via API with DroFx.pptx
ELKO dropshipping via API with DroFx.pptx
Ā 
BabyOno dropshipping via API with DroFx.pptx
BabyOno dropshipping via API with DroFx.pptxBabyOno dropshipping via API with DroFx.pptx
BabyOno dropshipping via API with DroFx.pptx
Ā 
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Vip Mumbai Call Girls Thane West Call On 9920725232 With Body to body massage...
Ā 
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
Ā 
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
Ā 
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Call Girls Indiranagar Just Call šŸ‘— 7737669865 šŸ‘— Top Class Call Girl Service B...
Ā 

Strata London - Deep Learning 05-2015

  • 1. Deep learning made doubly easy with reusable deep features Carlos Guestrin Dato, CEO University of Washington, Amazon Prof. of ML
  • 2. Successful apps in 2015 must be intelligent Machine learning key to next-gen apps ā€¢ Recommenders ā€¢ Fraud detection ā€¢ Ad targeting ā€¢ Financial models ā€¢ Personalized medicine ā€¢ Churn prediction ā€¢ Smart UX (video & text) ā€¢ Personal assistants ā€¢ IoT ā€¢ Socials nets ā€¢ ā€¦Last decade: Data management Now: Intelligent apps ? Last 5 years: Traditional analytics
  • 3. The ML pipeline circa 2013 DATA ML Algorithm My curve is better than your curve Write a paper
  • 4. 2015: Production ML pipeline DATA YourWebServiceor IntelligentApp ML Algorithm Data cleaning & feature eng Offline eval & Parameter search Deploy model Data engineering Data intelligence Deployment Using deep learning Goal: Platform to help implement, manage, optimize entire pipeline
  • 5. Todayā€™s talk Features in ML Neural networks Deep learning for computer vision Deep learning made easy with deep features Applications to text data Deployment in production
  • 6. Features are key to machine learning
  • 7. 7 Simple example: Spam filtering ā€¢ A user walks into an emailā€¦ - Will she thinks its spam? ā€¢ Whatā€™s the probability email is spam? Text of email User info Source info Input: x MODEL Yes! No Output: Probability of y
  • 8. 8 Feature engineering: the painful black art of transforming raw inputs into useful inputs for ML algorithm ā€¢ E.g., important words, stemming text, complex transformation of inputs,ā€¦ MODEL Yes! No Output: Probability of y Feature extraction Features: Ī¦(x) Text of email User info Source info Input: x
  • 10. 10 Linear classifiers ā€¢ Most common classifier - Logistic regression - SVMs - ā€¦ ā€¢ Decision correspond to hyperplane: - Line in high dimensional space w0 + w1 x1 + w2 x2 > 0 w0 + w1 x1 + w2 x2 < 0
  • 11. 11 Graph representation of classifier: useful for defining neural networks x 1 x 2 x d y ā€¦ 1 w2 w0 + w1 x1 + w2 x2 + ā€¦ + wd xd > 0, output 1 < 0, output 0 Input Output
  • 12. 12 What can a linear classifier represent x1 OR x2 x1 AND x2 x 1 x 2 1 y -0.5 1 1 x 1 x 2 1 y -1.5 1 1
  • 13. 13 What canā€™t a simple linear classifier represent? XOR the counterexample to everything Need non-linear features
  • 14. Solving the XOR problem: Adding a layer XOR = x1 AND NOT x2 OR NOT x1 AND x2 z 1 -0.5 1 -1 z1 z2 z 2 -0.5 -1 1 x 1 x 2 1 y 1 -0.5 1 1 Thresholded to 0 or 1
  • 15. 15 A neural network ā€¢ Layers and layers and layers of linear models and non-linear transformation ā€¢ Around for about 50 years - Fell in ā€œdisfavorā€ in 90s ā€¢ In last few years, big resurgence - Impressive accuracy on a several benchmark problems - Powered by huge datasets, GPUs, & modeling/learning alg improvements x 1 x 2 1 z 1 z 2 1 y
  • 16. Applications to computer vision (or the deep devil is in the deep details)
  • 17. 17 Image features ā€¢ Features = local detectors - Combined to make prediction - (in reality, features are more low-level) Face! Eye Eye Nose Mouth
  • 18. 18 Many hand create features existā€¦ Computer$vision$features$ SIFT$ Spin$image$ HoG$ RIFT$ Textons$ GLOH$ Slide$Credit:$Honglak$Lee$
  • 19. 19 Standard image classification approach Input Computer$vision$features$ SIFT$ Spin$image$ HoG$ RIFT$ Textons$ GLOH$ Slide$Credit:$Honglak$Lee$ Extract features Use simple classifier e.g., logistic regression, SVMs Car?
  • 20. 20 Many hand create features existā€¦ Computer$vision$features$ SIFT$ Spin$image$ HoG$ RIFT$ Textons$ GLOH$ Slide$Credit:$Honglak$Lee$ ā€¦ but very painful to design
  • 21. 21 Use neural network to learn features Each layer learns features, at different levels of abstraction Y LeCun MA Ranzato Deep Learning = Learning Hierarchical Representations It's deep if it has more than one stage of non-linear feature transformation Trainable Classifier Low-Level Feature Mid-Level Feature High-Level Feature Feature visualization of convolutional net trained on ImageNet from [ Zeiler & Fergus 2013]
  • 22. 22 Many tricks needed to work wellā€¦ ā€¢ Different types of layers, connections,ā€¦ needed for high accuracy Krizhevsky et al. ā€˜12
  • 24. Sample results ā€¢ Traffic sign recognition (GTSRB) - 99.2% accuracy ā€¢ House number recognition (Google) - 94.3% accuracy 30
  • 25. Krizhevsky et al. ā€™12: 60M parameters, won 2012 ImageNet competition 31
  • 26. 32 ImageNet 2012 competition: 1.5M images, 1000 categories 32
  • 28. 34 Application to scene parsing Ā©Carlos Guestrin 2005-2014 Y LeCun MA Ranzato Semantic Labeling: Labeling every pixel with the object it belongs to [ Farabet et al. ICML 2012, PAMI 2013] Would help identify obstacles, targets, landing sites, dangerous areas Would help line up depth map with edge maps
  • 29. Challenges of deep learning
  • 30. Deep learning score card Pros ā€¢ Enables learning of features rather than hand tuning ā€¢ Impressive performance gains on - Computer vision - Speech recognition - Some text analysis ā€¢ Potential for much more impact Cons
  • 31. Deep learning workflow Lots of labeled data Training set Validation set 80% 20% Learn deep neural net model Validate
  • 32. Deep learning score card Pros ā€¢ Enables learning of features rather than hand tuning ā€¢ Impressive performance gains on - Computer vision - Speech recognition - Some text analysis ā€¢ Potential for much more impact Cons ā€¢ Computationally really expensive ā€¢ Requires a lot of data for high accuracy ā€¢ Extremely hard to tune - Choice of architecture - Parameter types - Hyperparameters - Learning algorithm - ā€¦ ā€¢ Computational + so many choices = incredibly hard to tune
  • 34. 40 Change image classification approach? Input Computer$vision$features$ SIFT$ Spin$image$ HoG$ RIFT$ Textons$ GLOH$ Slide$Credit:$Honglak$Lee$ Extract features Use simple classifier e.g., logistic regression, SVMs Car? Can we learn features from data, even when we donā€™t have data or time?
  • 35. 41 Transfer learning: Use data from one domain to help learn on another Lots of data: Learn neural net Great accuracy on cat v. dogvs. Some data: Neural net as feature extractor + Simple classifier Great accuracy on 101 categories Old idea, explored for deep learning by Donahue et al. ā€™14
  • 36. 42 Whatā€™s learned in a neural net Neural net trained for Task 1: cat vs. dog Very specific to Task 1 Should be ignored for other tasks More generic Can be used as feature extractor vs.
  • 37. 43 Transfer learning in more detailā€¦ Neural net trained for Task 1: cat vs. dog Very specific to Task 1 Should be ignored for other tasks More generic Can be used as feature extractor Keep weights fixed! For Task 2, predicting 101 categories, learn only end part Use simple classifier e.g., logistic regression, SVMs Class ?
  • 38. 44 Careful where you cutā€¦ Last few layers tend to be too specific Y LeCun MA Ranzato Deep Learning = Learning Hierarchical Representations It's deep if it has more than one stage of non-linear feature transformation Trainable Classifier Low-Level Feature Mid-Level Feature High-Level Feature Feature visualization of convolutional net trained on ImageNet from [ Zeiler & Fergus 2013] Too specific for car detectionUse these!
  • 39. Transfer learning with deep features Training set Validation set 80% 20% Learn simple model Some labeled data Extract features with neural net trained on different task Validate Deploy in production
  • 40. How general are deep features?
  • 42. Simple text classification with bag of words aardvark 0 about 2 all 2 Africa 1 apple 0 anxious 0 ... gas 1 ... oil 1 ā€¦ Zaire 0 Use simple classifier e.g., logistic regression, SVMs Class ? One ā€œfeatureā€ per word
  • 43. Word2Vec: Neural network for finding high dimensional representation per word Mikolov et al. ā€˜13 Skip-gram Model: From a word, predict nearby words in sentence Awesome learning talk at Strata deep 300 dim representation 300 dim representation 300 dim representation 300 dim representation 300 dim representation 300 dim representation Neural net Viewed as deep features
  • 44. 50 Related words placed nearby high dim space Projecting 300 dim space into 2 dim with PCA (Mikolov et al. ā€™13)
  • 45. Classifier: e.g., logistic regression, SVMs with 300 x number_of_words parameters Class ? Embed each word into 300 dim space Text classification with word embeddings aardvark 0 about 2 all 2 Africa 1 apple 0 anxious 0 ... gas 1 ... oil 1 ā€¦ Zaire 0
  • 48. Deploying ML in production
  • 49. 55 DATA ML Algorithm Deployment? ā€¢ Write spec, other team implements in ā€˜productionā€™ language o 6-12 months o Stale/irrelevant model/approach o 2 teams maintaining 2 systems Custom Model Data Engineers, Data Architects, DevOps, App Developers App A P I Data Scientist
  • 50. ML deployment requirements 56 Easy to integrate Rest API Scalable Fault tolerant Flexible Any model, any Python App A P I A P I C A C H E A P I C A C H E A P I C A C H E LB GLC Model GLC Model GLC Model Dato Models Dato Models Dato Models A P I C A C H E A P I C A C H E A P I C A C H E LB GLC Model GLC Model GLC Model Python Python Python
  • 51. 57 Do-It-Yourself ā€¢ Web Service layer: - Tornado, Flask, Keen, Django, ā€¦ ā€¢ Caching layer: - Redis, Cassandra, Memcached, DynamoDb, MySQL, ā€¦ ā€¢ Logs: - Logback, LogStash, Splunk, Loggly, ā€¦ ā€¢ Metrics: - AWS CloudWatch, Mixpanel, Librato, ā€¦ A P I C A C H E A P I C A C H E A P I C A C H E LB GLC Model GLC Model GLC Model Python Python Python App
  • 52. 58 ā€¦ or use Dato Predictive Services YourWebServiceor IntelligentApp ML Model Dato Predictive Services CachingLayer Predictive ObjectServer Serves predictions in a robust, scalable, incremental fashion Better ML Model Serve any model: GraphLab Create, scikit-learn, Python, ā€¦
  • 53. ā€¢ Out-of-core computation ā€¢ Tools for feature engineering ā€¢ Rich data type support ā€¢ Models built for scale ā€¢ App-oriented toolkits ā€¢ Advanced ML & Extensible ā€¢ Deploy models as low-latency REST services ā€¢ Same code for distributed computation ā€¢ Elastically scale up or out with one command ā€¢ Job monitoring & model management ā€¢ Deploy existing Python code & models ā€¢ Run on AWS EC2 or Hadoop Yarn SGraph Create Engine SFrameCanvas Machine Learning Toolkits SDK GraphLab Create Dato Distributed Dato Predictive Services Predictive Engine REST Client Direct Model Mgmt Distributed Engine DirectJob Client Job Mgmt Dato Platform
  • 55. Deep learning made easy with deep features Deep learning: exciting ML development Slow, lots of tuning, needs lots of data Deep features: reuse deep models for new domains Needs less data Faster training times Much simpler tuning Can still achieve excellent performance

Editor's Notes

  1. So I got started with ML by taking a class. Data -> to ML algo, and then generate a plot. Of course this isnā€™t how actual applications are written, but this is often where customers are starting when approaching taking ML to production.