This document discusses the evolution and impact of deep learning in artificial intelligence, highlighting key figures and institutions involved in its development. It reviews the advantages of using unlabeled data to improve model performance and emphasizes the significance of feature learning and deep architectures. Additionally, it covers various applications of deep learning across fields such as speech recognition, image classification, and natural language processing.

1. Deep Learning
for Data Scientists
Andrew B. Gardner
agardner@momentics.com
http://linkd.in/1byADxC
www.momentics.com/deep-learning

3. Deep Learning in the Press…
Ng
Hinton
LeCun
Zuckerberg
Google Hires Brains that Helped
Supercharge Machine Learning.
Wired 3/2013.
Kurzweil
Facebook taps ‘Deep Learning’ Giant for New AI Lab.
Wired 12/2013.
Is “Deep Learning” A Revolutions in
Artificial Intelligence?
The Man Behind the Google Brain: Andrew
Ng and the Quest for the New AI.
New Yorker 11/2012.
Wired 5/2013.
New Techniques from Google and Ray Kurzweil Are
Taking Artificial Intelligence to Another Level.
MIT Technology Review 5/2013.

4. … Publication & Search Trends …
Google Scholar Citations
Google Trends
600
big data
500
data science
400
300
“deep learning” +
“neural network”
deep learning
machine learning
200
100
0
‘06
‘11
‘06
‘11
domains: computer vision, speech & audio, bioinformatics, etc.
Conferences: NIPS, ICLR, ICML, …

5. … Industry & Products
• Google
Microsoft Real-time English-Chinese Translation
– Android Voice
Recognition
– Maps
– Image+
•
•
•
•
SIRI
Translation
Documents
…
https://www.youtube.com/watch?v=Nu-nlQqFCKg
Microsoft Chief Research Officer Rick
Rashid, 11/2012

6. Deep Learning Epicenters (North America)
de Freitas (UBC)
Microsoft
Bengio (U Montreal)
Hinton (U Toronto)
Facebook
Ng (Stanford)
Google
Yahoo
LeCun (NYU)

7. Deep Learning: The Origin Story

8. Before: A Cat Detector
We want to build this….
classifier
f : X ®Y
Y ~ the labels {“cat”, “dog”}
X ~ the images
… for less than $1.0M !

9. Challenge: Labeled Data
Labels are expensive  Less data
Intuitively: more data is good
cat
cat
dog
unused,unlabeled
cat
dog

10. Challenge: Features
Features are expensive  Fewer, shallow
Intuitively: better features are good
image (pixels)
Magic feature dictionary
SIFT
HoG
B W
SIFT
binary histogram
Moments
Shape Histogram
+
++
+
+++
+
+
+
+ +
+
x=(1.3, 2.8, …)
Fang detector
Something new

11. Machine Learning (Before)
Building a Cat Detector 1.0
expensive
important*
Features
Detector
(Classifier)

12. fa
ng
of
in
ch
on,
of
of
us
on
is
is
bly
How Good is “More Data?”
speech. The memory-based learner used only
the word before and word after as features.
Labels are expensive  Less data
1.00
• More data dominates*
better techniques
.975
0.95
0.90
Test Accuracy
a
93,
In
is
fic
es
are
m
ber
• Often have lots of data
0.85
.825
0.80
Memory-Based
Winnow
0.75
Perceptron
Naïve Bayes
0.70
0.1
1
10
100
Millions of Words
1000
Learning curves for confusion set
Figure 1. Learning Curves for Confusion Set
disambiguation, e.g. {to, two, too}.
Disambiguation
We collected a 1
-billion-word training
corpus from a variety of English texts, including
• … we just don’t have
lots of labels
• What if there was a
way to use unlabeled
data?
“Scaling to Very Very Large Corpora for Natural Language Disambiguation,” Banko and Brill, 2001.

13. The Impact of Features
Intuitively: better features are good
• Critical to success – even more than data!
• How to create / engineer features?
– Typically shallow
• Domain-specific
• What if there was a way to automatically
learn features?

14. Machine Learning (What We Want)
Building a Cat Detector 2.0
bountiful
important*
Features + Detector
(Classifier)
end-to-end

15. AR” Building an Object Recognition System
”
“CAR”
Deep Nets Intuition
“CAR”
car
intermediate representations
CLASSIFIER
FEATURE
EXTRACTOR
label
IDEA: Use data to optimize features for the given task.
olutional DBN's for scalable unsup. learning...” ICML 2009
Lee et al. ICML 2009
12
Ranzato
2
Ranzato
13
Ranzato
Ranza

16. on from low
structure as
hical Another Example of Hierarchy
Learning
rchical Learning
mplexity from low
progression
ral progression from low
high level structure as
to high level structure as
natural complexity
in natural complexity
what is being
eto monitor whatisisbeing
the machine
o monitor what being
r
and guide the machine
es toto guide themachine
t and
er subspaces
tter subspaces
od lower level
llower level heads
ntation can be used for
sentation can be usedfor
ndistinct tasks for
be used
istinct tasks
s
faces
as
parts
edges

17. d tomachine machine
e guide the
he
subspaces Hierarchy Reusability?
faces
cars
elephants
chairs
wer level
be used forbe used for
tation can
tinct tasks
5
5

18. A Breakthrough
G. E. Hinton, S. Osindero, and Y. Teh, “A fast learning
algorithm for deep belief nets,” Neural
Computation, vol. 18, pp. 1527–1554, 2006.
G. E. Hinton and R. R. Salakhutdniov, “Reducing the
dimensionality of data with neural networks,”
Science, vol. 313, no. 5786, pp. 504-507, July 2006.
before
after

19. Deep Belief Nets
MNIST
60K + 10K Images
Technique
Test Error
DBN pretrain
1.25
SVM
1.4
kNN
2.8-4.4
ConvNet
0.4 -> 0.23
supervised tuning
unsupervised pretraining

20. MNIST Sample Errors
Ciresan et al. “Deep Big Simple Neural Networks Excel on
Handwritten Digit Recognition,” 2010

21. Key Ideas
• Learn features from data
– Use all data
• Deep architecture
– Representation
– Computational efficiency
– Shared statistics
• Practical training
• State-of-the-art (it worked)

22. After: Cat Detector
unlabeled images (millions)
labeled images (few)
deep learning
network
more data
automatic (deep) features

23. How Does It Work?

24. This Is A Neuron
output
1. Sum all inputs (weighted)
y
x = w0 + w1z1 + w2 z2 + w3z3
f(x)
2. Nonlinearly transform
y = f ( x)
weights
w0 w1
w2
sigmoid
w3
tanh
1
bias
z1
z2
inputs
z3
activation function

25. A Neural Network
forward propagation: weighted sum inputs, produce activation, feed forward
cat
dog
Output
Hidden
13.5
weight
21
n_teeth
16
n_whiskers
Inputs
(the features)

26. Training
Back propagation of error.
1
0
cat
dog
total error at top
proportional
contributions going
backwards
13.5
weight
21
n_teeth
16
n_whiskers

27. After Training
network
layer weights
weights as a matrix
[.5, -.2, 4, .15, -1,…]
-.5
.4
0
.1
.1
.5
-1
2
[-.5, -.3, .4, 0, …]
-.3
.7
-.2
.4
we can view weight matrix as image
… plus performance evaluation & logging

28. Building Blocks
So many choices!
network topology
• Network Topology
– Number of layers
– Nodes per layer
• Layer Type
– Feedforward
– Restricted Boltzmann
– Autoencoder
– Recurrent
– Convolutional
layer type
neuron type
• Neuron Type
– Rectified Linear Unit
• Regularization
– Dropout
• Magic Numbers

29. A Deep Learning Recipe, 1.0
• Lots of data, some+
labels
• Train each RBM layer
greedily, successively
• Add an output layer
and train with labels
labels

30. A Few Other Important Things
• Deep Learning Recipe 2.0
– Dropout / regularization
– Rectified Linear Units
•
•
•
•
Convolutional networks
Hyperparameters
Not just neural networks
Practical Issues (GPU)

31. Some Applications

32. Sample Classification Results
ImageNet
V
alidation classification
Krizhevsky et al., NIPS 2012.
[Krizhevsky et al. NI PS’12

33. Segmentation
neuronal membranes
Ciresan et al. “DNN segment neuronal membranes...” NIPS 2012

34. CalTech 256 2 5 6
Caltech
Z eiler & Fergus, Vis
ualizing and Unders
tanding Convolutional Ne
tworks arXiv 1311.2901, 2013
,
7
5
7
0
6
5
6 training examples
6
0
5
5
5
0
4
5
4
0
3
5
3
0
2
5
0
1
0
2
0
3
0
4
0
5
0
6
0
Zeiler & Fergus,”Visualizing and Understanding Convolutional Networks,” arXiv 1311.2901, 2013

35. Application: Speech
frequencies
in window
“He can for example present significant university wide
issues to the senate.”
small time window
slide 15ms
phoneme
Spectrogram: window in time -> vector of frequences; slide; repeat

36. Automatic Speech
CDBNs for speech
Unlabeled TIMIT data -> convolutional DBN
Trained on unlabeled TIMIT corpus
Experimental R
• Speaker identification
TIMIT Speaker identification
Accuracy
Prior art (Reynolds, 1995)
99.7%
Convolutional DBN
100.0%
• Phone classification
TIMIT Phone classification
Accuracy
Clarkson et al. (1999)
77.6%
Gunawardana et al. (2005)
78.3%
Sung et al. (2007)
78.5%
Petrov et al. (2007)
78.6%
Sha & Saul (2006)
78.9%
Yu et al. (2009)
79.2%
Convolutional DBN
80.3%
Learned first-layer bases
Lee et al., “Unsupervised feature learning for audio classification using convolutional deep
68
belief networks”, NIPS 2009.

37. A Long List of Others
• Kaggle
– Merck Molecular Activation (‘12)
– Salary Prediction (‘13)
•
•
•
•
Learning to Play Atari Games (‘13)
NLP – chunking, NER, parsing, etc.
Activity recognition from video
Recommendations

38. Deep Learning In A Nutshell
•
•
•
•
•
•
•
•
Architectures vs. features
Deep vs. shallow
Automatic* features
Lots of data vs. best technique
Compute- vs. human intensive
State-of-the-art
Breaks expert, domain barrier
Details & tricks can be complex
http://www.deeplearning.net/

39. Interested in Deep Learning?
Connect for:
• Training Workshop (interest list)
• Projects / consulting
• Collaboration
• Questions
agardner@momentics.com
http://www.momentics.com/deep-learning/