SlideShare a Scribd company logo
1 of 35
Download to read offline
Natalino Busa
Head of Applied Data Science
Data Scientist, Big and Fast Data Architect
Currently at Teradata
Previously:
Enterprise Data Architect at ING
Senior Researcher at Philips Research
Interests:
Spark, Flink, Cassandra, Akka, Kafka, Mesos
Anomaly Detection, Time Series, Deep Learning
Data Science: approaches
Supervised:
- you know what the outcome must be
Unsupervised:
- you don’t know what the outcome must be
Semi-Supervised:
- You know the outcome only for some samples
Popularity of Neural Networks: “The cat neuron”
Andrew Ng, Jeff Dean et al:
1000 Machines
10 Million images
1 Billion connections
Train for 3 days
http://research.google.com/archive/unsupervised_icml2012.html
Popularity of Neural Networks: “AI at facebook”
Yann LeCunn
Director of AI research at Facebook
Ask the AI what it sees in the image
“Is there a baby?”
Facebook’s AI: “Yes.”
“What is the man doing?”
Facebook’s AI: “Typing.”
“Is the baby sitting on his lap?”
Facebook’s AI: “Yes.”
http://www.wired.com/2015/11/heres-how-smart-facebooks-ai-has-become/
Data Science: approaches
Supervised:
- you know what the outcome must be
Unsupervised:
- you don’t know what the outcome must be
Semi-Supervised:
- You know the outcome only for some samples
Unsupervised Learning
- Clustering, Feature extraction
Imagining, Medical data, Genetics, Crime patterns,
Recommender systems, Climate hot spots analysis, anomaly detection
… Given a set of items,
it answers the question “how can we efficiently describe the collection?
It defines a measure of “similarity” between items.
Supervised Learning
- Classification
Marketing Churn, Credit Loan, Success rate
Insurance Defaulting, Health conditions and patologies
Categorization of wine, real estates,
… Given the values of some properties,
it answers the question “to which class/group does this item belong?”
Classification: Dimensionality matters
- Number of dimensions or features of your input data
- Statistical relations, smoothness of the data
- Embedded space
input : 784 dimensions
output: 10 classes
input : 4 dimensions
output: 3 classes
28x28 pixels
AI, complexity and models
Does it do well on
Training Data ?
Does it do well on
Test Data ?
Bigger Neural Network
(rocket engine)
More Data
(rocket fuel)
yes yes
no
no
Done?
Different
Architecture
(new rocket)
no
https://www.youtube.com/watch?v=CLDisFuDnog
Evolution of Machine Learning
Input
Hand Designed
Program
Rule-based System
Output
Prof. Yoshua Bengio - Deep Learning
https://youtu.be/15h6MeikZNg
Evolution of Machine Learning
Input
Hand Designed
Program
Input
Rule-based System
Output
Hand Designed
Features
Mapping from
features
Output
Classic Machine
Learning
Prof. Yoshua Bengio - Deep Learning
https://youtu.be/15h6MeikZNg
Evolution of Machine Learning
Input
Hand Designed
Program
Input Input
Rule-based System
Output
Hand Designed
Features
Mapping from
features
Output
Learned
Features
Mapping from
features
Output
Classic Machine
Learning
Representational
Machine Learning
Prof. Yoshua Bengio - Deep Learning
https://youtu.be/15h6MeikZNg
Evolution of Machine Learning
Input
Hand Designed
Program
Input Input
Rule-based System
Output
Hand Designed
Features
Mapping from
features
Output
Learned
Features
Mapping from
features
Output
Classic Machine
Learning
Input
Learned
Features
Learned
Complex features
Output
Mapping from
features
Representational
Machine Learning
Deep Learning
Prof. Yoshua Bengio - Deep Learning
https://youtu.be/15h6MeikZNg
“dendrites”
Axon’s
response
Activation function
From Biology to a Mathematical Model
Logit model: Perceptron
1 Layer Neural Network
Takes: n-input features: Map them to a soft “binary” space
∑
x1
x2
xn
f
Multiple classes: Softmax
From soft binary space to predicting probabilities:
Take n inputs, Divide by the sum of the predicted values
∑
x1
x2
xn
f
∑ f
softmax Cat: 95%
Dog: 5% Values between 0 and 1
Sum of all outcomes = 1
It behaves like a probability,
But it’s just an estimate!
Cost function: Supervised Learning
The actual outcome is different than the desired outcome
We measure the difference!
This measure can be done in various ways:
- Mean absolute error (MAE)
- Mean squared error (MSE)
- Categorical Cross-Entropy
Compares estimated probability vs actual probability
Minimize cost: How to Learn?
The cost function depends on:
- Parameters of the model
- How the model “composes”
Goal :
modify the parameters to reduce the error!
Vintage math from last century
Build deeper networks
Stack layers of perceptrons
- “Sequential Network”
- Back propagate the error
SOFTMAX
Input parameters
Classes (estimated probabilities)
Feed-forward
Cost function
supervised : actual output
Correct
parameters
Some problems
- Calculating the derivative of the Cost function
- can be error prone
- Automation would be nice!
- Complex network graph = complex derivative
- Dense Layers (Fully connected)
- Harder to converge
- Number of parameters grows fast!
- Overfitting and Parsimony
- Learn “well”, generalization capacity
- Be efficient in the number of parameters
Some Solutions
- Calculating the derivative of the Cost function
- Software libraries
- GPU support for computing vectorial and tensorial data
- New Layers Types
- Convolution Layers 2D/3D
- Dropout layer
- Fast activation functions
- Faster learning methods
- Derived from Stochastic Gradient Descend (SGA)
- Weight initializations with Auto-Encoders and RBM
Convolutional Networks
Idea 1: reuse the weights across while scanning the image
Idea 2: subsampling results from layers to layers
Fast Activation Functions
Idea: don’t use complex exponential functions,
linear functions are fast to compute, and easy to differentiate !
Dropout Layer, Batch Weight Normalization
Dropout:
Set randomly some of the input to zero.
It improves generalization and makes the network function more robust to errors.
Batch Weight Normalization:
Normalize the activations of the previous layer at each batch.
Efficient Symbolic Differentiation
There are good libraries which calculate the derivatives symbolically of an
arbitrary number of stacked layers
● efficient symbolic differentiation
● dynamic C code generation
● transparent use of a GPU
CNTK
Efficient Symbolic Differentiation (2)
There are good libraries which calculate the derivatives symbolically of an
arbitrary number of stacked layers
● efficient symbolic differentiation
● dynamic C code generation
● transparent use of a GPU
>>> import theano
>>> import theano.tensor as T
>>> from theano import pp
>>> x = T.dscalar('x')
>>> y = x ** 2
>>> gy = T.grad(y, x)
>>> f = theano.function([x], gy)
pp(f.maker.fgraph.outputs[0])
'(2.0 * x)'
Higher Abstraction Layer: Keras
Keras: Deep Learning library for Theano and TensorFlow
- Easier to stack layers
- Easier to train and test
- More ready-made blocks
http://keras.io/
Example 1: Iris classification
Categorize Iris flowers based on
- Sepal length/width
- Petal length/width
3 classes,
Dataset is quite small (150 samples)
- Iris Setosa
- Iris Versicolour
- Iris Virginica
input : 4 dimensions
output: 3 classes
Iris classification: Network
model = Sequential()
model.add(Dense(15, input_shape=(4,)))
model.add(Activation('relu'))
model.add(Dropout(0.1))
model.add(Dense(10))
model.add(Activation('relu'))
model.add(Dropout(0.1))
model.add(Dense(nb_classes))
model.add(Activation('softmax'))
SOFTMAX
RELU
RELU
Setosa Versicolour Virginica
Dropout 10%
Dropout 10%
Train- Test split 80% - 20%
Test accuracy: 96%
Example 2: telecom customer marketing
Semi-synthetic dataset
The "churn" data set was developed to predict telecom customer churn based on information about their account. The data files state that the data are "artificial
based on claims similar to real world". These data are also contained in the C50 R package.
1 classes (churn)
Dataset is quite small (about 3000 samples)
17 input dimensions:
State, account length, area code, phone number,international plan,voice mail plan,number vmail messages,total day
minutes,total day calls,total day charge,total eve minutes,total eve calls,total eve charge,total night minutes,total night
calls,total night charge,total intl minutes,total intl calls,total intl charge,number customer service calls
Churn telecom: Network
model = Sequential()
model.add(Dense(50, input_shape=(17,)))
model.add(Activation("hard_sigmoid"))
model.add(BatchNormalization())
model.add(Dropout(0.1))
model.add(Dense(10))
model.add(Activation("hard_sigmoid"))
model.add(BatchNormalization())
model.add(Dropout(0.1))
model.add(Dense(1))
model.add(Activation(sigmoid))
SOFTMAX
RELU
RELU
Churn No-Churn
Dropout 10%
Dropout 10%
Train- Test split 80% - 20%
Test accuracy: 82%
Models: Small Data, Big Data
- Not all domains have large amount of data
- Think of Clinical Tests, or Lengthy/Costly Experimentations
- Small specialized data set and Neural Networks
- Good for complex non-linear separation of classes
Interesting Read:
https://medium.com/@ShaliniAnanda1/an-open-letter-to-yann-lecun-22b244fc0a5a#.ngpal1ojx
Conclusions
- Neural Networks can be used for small data as well
- Other methods might be more efficient in this scenario’s
- Neural Networks are an extension to GLMs and linear regression
- Learn Linear Regression, GLM, SVM as well
- Random Forests and Boosted Trees are an alternative
- More data = Bigger and better Neural Networks
- We have some tools to jump start analysis
Connect on Twitter and Linkedin !
Thanks!

More Related Content

Viewers also liked

Computing Professional Identity for the Economic Graph
Computing Professional Identity for the Economic GraphComputing Professional Identity for the Economic Graph
Computing Professional Identity for the Economic GraphVitaly Gordon
 
A survey on transfer learning
A survey on transfer learningA survey on transfer learning
A survey on transfer learningazuring
 
Best Blue Brain ppt ever.
Best Blue Brain ppt ever.Best Blue Brain ppt ever.
Best Blue Brain ppt ever.Suhail Shaikh
 
Beyond Matching: Applying Data Science Techniques to IOC-based Detection
Beyond Matching: Applying Data Science Techniques to IOC-based DetectionBeyond Matching: Applying Data Science Techniques to IOC-based Detection
Beyond Matching: Applying Data Science Techniques to IOC-based DetectionAlex Pinto
 
DELL project
DELL projectDELL project
DELL projectKIMEP
 
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...Alex Pinto
 
How to improve customer experience with a self organizing network
How to improve customer experience with a self organizing networkHow to improve customer experience with a self organizing network
How to improve customer experience with a self organizing networkComarch
 
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdA
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdANTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdA
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdANorth Texas Chapter of the ISSA
 
Deep Learning for Cyber Security
Deep Learning for Cyber SecurityDeep Learning for Cyber Security
Deep Learning for Cyber SecurityAltoros
 
User and entity behavior analytics: building an effective solution
User and entity behavior analytics: building an effective solutionUser and entity behavior analytics: building an effective solution
User and entity behavior analytics: building an effective solutionYolanta Beresna
 
NTXISSACSC4 - Hacking Performance Management, the Blue Green Game
NTXISSACSC4 - Hacking Performance Management, the Blue Green GameNTXISSACSC4 - Hacking Performance Management, the Blue Green Game
NTXISSACSC4 - Hacking Performance Management, the Blue Green GameNorth Texas Chapter of the ISSA
 
Artificial Neural Network Seminar - Google Brain
Artificial Neural Network Seminar - Google BrainArtificial Neural Network Seminar - Google Brain
Artificial Neural Network Seminar - Google BrainRawan Al-Omari
 
A very easy explanation to understanding machine learning (Supervised & Unsup...
A very easy explanation to understanding machine learning (Supervised & Unsup...A very easy explanation to understanding machine learning (Supervised & Unsup...
A very easy explanation to understanding machine learning (Supervised & Unsup...Ryo Onozuka
 
Scalable and Flexible Machine Learning With Scala @ LinkedIn
Scalable and Flexible Machine Learning With Scala @ LinkedInScalable and Flexible Machine Learning With Scala @ LinkedIn
Scalable and Flexible Machine Learning With Scala @ LinkedInVitaly Gordon
 
Biometric security using cryptography
Biometric security using cryptographyBiometric security using cryptography
Biometric security using cryptographySampat Patnaik
 

Viewers also liked (17)

Computing Professional Identity for the Economic Graph
Computing Professional Identity for the Economic GraphComputing Professional Identity for the Economic Graph
Computing Professional Identity for the Economic Graph
 
A survey on transfer learning
A survey on transfer learningA survey on transfer learning
A survey on transfer learning
 
Best Blue Brain ppt ever.
Best Blue Brain ppt ever.Best Blue Brain ppt ever.
Best Blue Brain ppt ever.
 
Beyond Matching: Applying Data Science Techniques to IOC-based Detection
Beyond Matching: Applying Data Science Techniques to IOC-based DetectionBeyond Matching: Applying Data Science Techniques to IOC-based Detection
Beyond Matching: Applying Data Science Techniques to IOC-based Detection
 
DELL project
DELL projectDELL project
DELL project
 
Ada boost
Ada boostAda boost
Ada boost
 
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...
Secure Because Math: A Deep-Dive on Machine Learning-Based Monitoring (#Secur...
 
How to improve customer experience with a self organizing network
How to improve customer experience with a self organizing networkHow to improve customer experience with a self organizing network
How to improve customer experience with a self organizing network
 
Docker で Deep Learning
Docker で Deep LearningDocker で Deep Learning
Docker で Deep Learning
 
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdA
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdANTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdA
NTXISSACSC4 - Identity as a Threat Plane Leveraging UEBA and IdA
 
Deep Learning for Cyber Security
Deep Learning for Cyber SecurityDeep Learning for Cyber Security
Deep Learning for Cyber Security
 
User and entity behavior analytics: building an effective solution
User and entity behavior analytics: building an effective solutionUser and entity behavior analytics: building an effective solution
User and entity behavior analytics: building an effective solution
 
NTXISSACSC4 - Hacking Performance Management, the Blue Green Game
NTXISSACSC4 - Hacking Performance Management, the Blue Green GameNTXISSACSC4 - Hacking Performance Management, the Blue Green Game
NTXISSACSC4 - Hacking Performance Management, the Blue Green Game
 
Artificial Neural Network Seminar - Google Brain
Artificial Neural Network Seminar - Google BrainArtificial Neural Network Seminar - Google Brain
Artificial Neural Network Seminar - Google Brain
 
A very easy explanation to understanding machine learning (Supervised & Unsup...
A very easy explanation to understanding machine learning (Supervised & Unsup...A very easy explanation to understanding machine learning (Supervised & Unsup...
A very easy explanation to understanding machine learning (Supervised & Unsup...
 
Scalable and Flexible Machine Learning With Scala @ LinkedIn
Scalable and Flexible Machine Learning With Scala @ LinkedInScalable and Flexible Machine Learning With Scala @ LinkedIn
Scalable and Flexible Machine Learning With Scala @ LinkedIn
 
Biometric security using cryptography
Biometric security using cryptographyBiometric security using cryptography
Biometric security using cryptography
 

More from Natalino Busa

Data Production Pipelines: Legacy, practices, and innovation
Data Production Pipelines: Legacy, practices, and innovationData Production Pipelines: Legacy, practices, and innovation
Data Production Pipelines: Legacy, practices, and innovationNatalino Busa
 
Data science apps powered by Jupyter Notebooks
Data science apps powered by Jupyter NotebooksData science apps powered by Jupyter Notebooks
Data science apps powered by Jupyter NotebooksNatalino Busa
 
7 steps for highly effective deep neural networks
7 steps for highly effective deep neural networks7 steps for highly effective deep neural networks
7 steps for highly effective deep neural networksNatalino Busa
 
Data science apps: beyond notebooks
Data science apps: beyond notebooksData science apps: beyond notebooks
Data science apps: beyond notebooksNatalino Busa
 
[Ai in finance] AI in regulatory compliance, risk management, and auditing
[Ai in finance] AI in regulatory compliance, risk management, and auditing[Ai in finance] AI in regulatory compliance, risk management, and auditing
[Ai in finance] AI in regulatory compliance, risk management, and auditingNatalino Busa
 
Strata London 16: sightseeing, venues, and friends
Strata  London 16: sightseeing, venues, and friendsStrata  London 16: sightseeing, venues, and friends
Strata London 16: sightseeing, venues, and friendsNatalino Busa
 
Real-Time Anomaly Detection with Spark MLlib, Akka and Cassandra
Real-Time Anomaly Detection  with Spark MLlib, Akka and  CassandraReal-Time Anomaly Detection  with Spark MLlib, Akka and  Cassandra
Real-Time Anomaly Detection with Spark MLlib, Akka and CassandraNatalino Busa
 
The evolution of data analytics
The evolution of data analyticsThe evolution of data analytics
The evolution of data analyticsNatalino Busa
 
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...Natalino Busa
 
Streaming Api Design with Akka, Scala and Spray
Streaming Api Design with Akka, Scala and SprayStreaming Api Design with Akka, Scala and Spray
Streaming Api Design with Akka, Scala and SprayNatalino Busa
 
Hadoop + Cassandra: Fast queries on data lakes, and wikipedia search tutorial.
Hadoop + Cassandra: Fast queries on data lakes, and  wikipedia search tutorial.Hadoop + Cassandra: Fast queries on data lakes, and  wikipedia search tutorial.
Hadoop + Cassandra: Fast queries on data lakes, and wikipedia search tutorial.Natalino Busa
 
Big data solutions for advanced marketing analytics
Big data solutions for advanced marketing analyticsBig data solutions for advanced marketing analytics
Big data solutions for advanced marketing analyticsNatalino Busa
 
Awesome Banking API's
Awesome Banking API'sAwesome Banking API's
Awesome Banking API'sNatalino Busa
 
Yo. big data. understanding data science in the era of big data.
Yo. big data. understanding data science in the era of big data.Yo. big data. understanding data science in the era of big data.
Yo. big data. understanding data science in the era of big data.Natalino Busa
 
Big and fast a quest for relevant and real-time analytics
Big and fast a quest for relevant and real-time analyticsBig and fast a quest for relevant and real-time analytics
Big and fast a quest for relevant and real-time analyticsNatalino Busa
 
Big Data and APIs - a recon tour on how to successfully do Big Data analytics
Big Data and APIs - a recon tour on how to successfully do Big Data analyticsBig Data and APIs - a recon tour on how to successfully do Big Data analytics
Big Data and APIs - a recon tour on how to successfully do Big Data analyticsNatalino Busa
 
Strata 2014: Data science and big data trending topics
Strata 2014: Data science and big data trending topicsStrata 2014: Data science and big data trending topics
Strata 2014: Data science and big data trending topicsNatalino Busa
 
Streaming computing: architectures, and tchnologies
Streaming computing: architectures, and tchnologiesStreaming computing: architectures, and tchnologies
Streaming computing: architectures, and tchnologiesNatalino Busa
 

More from Natalino Busa (20)

Data Production Pipelines: Legacy, practices, and innovation
Data Production Pipelines: Legacy, practices, and innovationData Production Pipelines: Legacy, practices, and innovation
Data Production Pipelines: Legacy, practices, and innovation
 
Data science apps powered by Jupyter Notebooks
Data science apps powered by Jupyter NotebooksData science apps powered by Jupyter Notebooks
Data science apps powered by Jupyter Notebooks
 
7 steps for highly effective deep neural networks
7 steps for highly effective deep neural networks7 steps for highly effective deep neural networks
7 steps for highly effective deep neural networks
 
Data science apps: beyond notebooks
Data science apps: beyond notebooksData science apps: beyond notebooks
Data science apps: beyond notebooks
 
[Ai in finance] AI in regulatory compliance, risk management, and auditing
[Ai in finance] AI in regulatory compliance, risk management, and auditing[Ai in finance] AI in regulatory compliance, risk management, and auditing
[Ai in finance] AI in regulatory compliance, risk management, and auditing
 
Strata London 16: sightseeing, venues, and friends
Strata  London 16: sightseeing, venues, and friendsStrata  London 16: sightseeing, venues, and friends
Strata London 16: sightseeing, venues, and friends
 
Data in Action
Data in ActionData in Action
Data in Action
 
Real-Time Anomaly Detection with Spark MLlib, Akka and Cassandra
Real-Time Anomaly Detection  with Spark MLlib, Akka and  CassandraReal-Time Anomaly Detection  with Spark MLlib, Akka and  Cassandra
Real-Time Anomaly Detection with Spark MLlib, Akka and Cassandra
 
The evolution of data analytics
The evolution of data analyticsThe evolution of data analytics
The evolution of data analytics
 
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...
Towards Real-Time banking API's: Introducing Coral, a web api for realtime st...
 
Streaming Api Design with Akka, Scala and Spray
Streaming Api Design with Akka, Scala and SprayStreaming Api Design with Akka, Scala and Spray
Streaming Api Design with Akka, Scala and Spray
 
Hadoop + Cassandra: Fast queries on data lakes, and wikipedia search tutorial.
Hadoop + Cassandra: Fast queries on data lakes, and  wikipedia search tutorial.Hadoop + Cassandra: Fast queries on data lakes, and  wikipedia search tutorial.
Hadoop + Cassandra: Fast queries on data lakes, and wikipedia search tutorial.
 
Big data solutions for advanced marketing analytics
Big data solutions for advanced marketing analyticsBig data solutions for advanced marketing analytics
Big data solutions for advanced marketing analytics
 
Awesome Banking API's
Awesome Banking API'sAwesome Banking API's
Awesome Banking API's
 
Yo. big data. understanding data science in the era of big data.
Yo. big data. understanding data science in the era of big data.Yo. big data. understanding data science in the era of big data.
Yo. big data. understanding data science in the era of big data.
 
Big and fast a quest for relevant and real-time analytics
Big and fast a quest for relevant and real-time analyticsBig and fast a quest for relevant and real-time analytics
Big and fast a quest for relevant and real-time analytics
 
Big Data and APIs - a recon tour on how to successfully do Big Data analytics
Big Data and APIs - a recon tour on how to successfully do Big Data analyticsBig Data and APIs - a recon tour on how to successfully do Big Data analytics
Big Data and APIs - a recon tour on how to successfully do Big Data analytics
 
Strata 2014: Data science and big data trending topics
Strata 2014: Data science and big data trending topicsStrata 2014: Data science and big data trending topics
Strata 2014: Data science and big data trending topics
 
Streaming computing: architectures, and tchnologies
Streaming computing: architectures, and tchnologiesStreaming computing: architectures, and tchnologies
Streaming computing: architectures, and tchnologies
 
Big data landscape
Big data landscapeBig data landscape
Big data landscape
 

Recently uploaded

Easter Eggs From Star Wars and in cars 1 and 2
Easter Eggs From Star Wars and in cars 1 and 2Easter Eggs From Star Wars and in cars 1 and 2
Easter Eggs From Star Wars and in cars 1 and 217djon017
 
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Jack DiGiovanna
 
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxNLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxBoston Institute of Analytics
 
RadioAdProWritingCinderellabyButleri.pdf
RadioAdProWritingCinderellabyButleri.pdfRadioAdProWritingCinderellabyButleri.pdf
RadioAdProWritingCinderellabyButleri.pdfgstagge
 
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...dajasot375
 
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...Boston Institute of Analytics
 
Generative AI for Social Good at Open Data Science East 2024
Generative AI for Social Good at Open Data Science East 2024Generative AI for Social Good at Open Data Science East 2024
Generative AI for Social Good at Open Data Science East 2024Colleen Farrelly
 
20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdfHuman37
 
Data Factory in Microsoft Fabric (MsBIP #82)
Data Factory in Microsoft Fabric (MsBIP #82)Data Factory in Microsoft Fabric (MsBIP #82)
Data Factory in Microsoft Fabric (MsBIP #82)Cathrine Wilhelmsen
 
INTERNSHIP ON PURBASHA COMPOSITE TEX LTD
INTERNSHIP ON PURBASHA COMPOSITE TEX LTDINTERNSHIP ON PURBASHA COMPOSITE TEX LTD
INTERNSHIP ON PURBASHA COMPOSITE TEX LTDRafezzaman
 
ASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel CanterASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel Cantervoginip
 
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样vhwb25kk
 
Advanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsAdvanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsVICTOR MAESTRE RAMIREZ
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFAAndrei Kaleshka
 
Heart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectHeart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectBoston Institute of Analytics
 
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024thyngster
 
While-For-loop in python used in college
While-For-loop in python used in collegeWhile-For-loop in python used in college
While-For-loop in python used in collegessuser7a7cd61
 
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort servicejennyeacort
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)jennyeacort
 

Recently uploaded (20)

Easter Eggs From Star Wars and in cars 1 and 2
Easter Eggs From Star Wars and in cars 1 and 2Easter Eggs From Star Wars and in cars 1 and 2
Easter Eggs From Star Wars and in cars 1 and 2
 
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
Building on a FAIRly Strong Foundation to Connect Academic Research to Transl...
 
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptxNLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
NLP Project PPT: Flipkart Product Reviews through NLP Data Science.pptx
 
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
Deep Generative Learning for All - The Gen AI Hype (Spring 2024)
 
RadioAdProWritingCinderellabyButleri.pdf
RadioAdProWritingCinderellabyButleri.pdfRadioAdProWritingCinderellabyButleri.pdf
RadioAdProWritingCinderellabyButleri.pdf
 
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
Indian Call Girls in Abu Dhabi O5286O24O8 Call Girls in Abu Dhabi By Independ...
 
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
NLP Data Science Project Presentation:Predicting Heart Disease with NLP Data ...
 
Generative AI for Social Good at Open Data Science East 2024
Generative AI for Social Good at Open Data Science East 2024Generative AI for Social Good at Open Data Science East 2024
Generative AI for Social Good at Open Data Science East 2024
 
20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf20240419 - Measurecamp Amsterdam - SAM.pdf
20240419 - Measurecamp Amsterdam - SAM.pdf
 
Data Factory in Microsoft Fabric (MsBIP #82)
Data Factory in Microsoft Fabric (MsBIP #82)Data Factory in Microsoft Fabric (MsBIP #82)
Data Factory in Microsoft Fabric (MsBIP #82)
 
INTERNSHIP ON PURBASHA COMPOSITE TEX LTD
INTERNSHIP ON PURBASHA COMPOSITE TEX LTDINTERNSHIP ON PURBASHA COMPOSITE TEX LTD
INTERNSHIP ON PURBASHA COMPOSITE TEX LTD
 
ASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel CanterASML's Taxonomy Adventure by Daniel Canter
ASML's Taxonomy Adventure by Daniel Canter
 
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样
1:1定制(UQ毕业证)昆士兰大学毕业证成绩单修改留信学历认证原版一模一样
 
Advanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsAdvanced Machine Learning for Business Professionals
Advanced Machine Learning for Business Professionals
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFA
 
Heart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectHeart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis Project
 
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024
Consent & Privacy Signals on Google *Pixels* - MeasureCamp Amsterdam 2024
 
While-For-loop in python used in college
While-For-loop in python used in collegeWhile-For-loop in python used in college
While-For-loop in python used in college
 
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
9711147426✨Call In girls Gurgaon Sector 31. SCO 25 escort service
 
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
Call Us ➥97111√47426🤳Call Girls in Aerocity (Delhi NCR)
 

Deep learning overview and practical use in marketing and cyber-security

  • 1. Natalino Busa Head of Applied Data Science
  • 2. Data Scientist, Big and Fast Data Architect Currently at Teradata Previously: Enterprise Data Architect at ING Senior Researcher at Philips Research Interests: Spark, Flink, Cassandra, Akka, Kafka, Mesos Anomaly Detection, Time Series, Deep Learning
  • 3. Data Science: approaches Supervised: - you know what the outcome must be Unsupervised: - you don’t know what the outcome must be Semi-Supervised: - You know the outcome only for some samples
  • 4. Popularity of Neural Networks: “The cat neuron” Andrew Ng, Jeff Dean et al: 1000 Machines 10 Million images 1 Billion connections Train for 3 days http://research.google.com/archive/unsupervised_icml2012.html
  • 5. Popularity of Neural Networks: “AI at facebook” Yann LeCunn Director of AI research at Facebook Ask the AI what it sees in the image “Is there a baby?” Facebook’s AI: “Yes.” “What is the man doing?” Facebook’s AI: “Typing.” “Is the baby sitting on his lap?” Facebook’s AI: “Yes.” http://www.wired.com/2015/11/heres-how-smart-facebooks-ai-has-become/
  • 6. Data Science: approaches Supervised: - you know what the outcome must be Unsupervised: - you don’t know what the outcome must be Semi-Supervised: - You know the outcome only for some samples
  • 7. Unsupervised Learning - Clustering, Feature extraction Imagining, Medical data, Genetics, Crime patterns, Recommender systems, Climate hot spots analysis, anomaly detection … Given a set of items, it answers the question “how can we efficiently describe the collection? It defines a measure of “similarity” between items.
  • 8. Supervised Learning - Classification Marketing Churn, Credit Loan, Success rate Insurance Defaulting, Health conditions and patologies Categorization of wine, real estates, … Given the values of some properties, it answers the question “to which class/group does this item belong?”
  • 9. Classification: Dimensionality matters - Number of dimensions or features of your input data - Statistical relations, smoothness of the data - Embedded space input : 784 dimensions output: 10 classes input : 4 dimensions output: 3 classes 28x28 pixels
  • 10. AI, complexity and models Does it do well on Training Data ? Does it do well on Test Data ? Bigger Neural Network (rocket engine) More Data (rocket fuel) yes yes no no Done? Different Architecture (new rocket) no https://www.youtube.com/watch?v=CLDisFuDnog
  • 11. Evolution of Machine Learning Input Hand Designed Program Rule-based System Output Prof. Yoshua Bengio - Deep Learning https://youtu.be/15h6MeikZNg
  • 12. Evolution of Machine Learning Input Hand Designed Program Input Rule-based System Output Hand Designed Features Mapping from features Output Classic Machine Learning Prof. Yoshua Bengio - Deep Learning https://youtu.be/15h6MeikZNg
  • 13. Evolution of Machine Learning Input Hand Designed Program Input Input Rule-based System Output Hand Designed Features Mapping from features Output Learned Features Mapping from features Output Classic Machine Learning Representational Machine Learning Prof. Yoshua Bengio - Deep Learning https://youtu.be/15h6MeikZNg
  • 14. Evolution of Machine Learning Input Hand Designed Program Input Input Rule-based System Output Hand Designed Features Mapping from features Output Learned Features Mapping from features Output Classic Machine Learning Input Learned Features Learned Complex features Output Mapping from features Representational Machine Learning Deep Learning Prof. Yoshua Bengio - Deep Learning https://youtu.be/15h6MeikZNg
  • 16. Logit model: Perceptron 1 Layer Neural Network Takes: n-input features: Map them to a soft “binary” space ∑ x1 x2 xn f
  • 17. Multiple classes: Softmax From soft binary space to predicting probabilities: Take n inputs, Divide by the sum of the predicted values ∑ x1 x2 xn f ∑ f softmax Cat: 95% Dog: 5% Values between 0 and 1 Sum of all outcomes = 1 It behaves like a probability, But it’s just an estimate!
  • 18. Cost function: Supervised Learning The actual outcome is different than the desired outcome We measure the difference! This measure can be done in various ways: - Mean absolute error (MAE) - Mean squared error (MSE) - Categorical Cross-Entropy Compares estimated probability vs actual probability
  • 19. Minimize cost: How to Learn? The cost function depends on: - Parameters of the model - How the model “composes” Goal : modify the parameters to reduce the error! Vintage math from last century
  • 20. Build deeper networks Stack layers of perceptrons - “Sequential Network” - Back propagate the error SOFTMAX Input parameters Classes (estimated probabilities) Feed-forward Cost function supervised : actual output Correct parameters
  • 21. Some problems - Calculating the derivative of the Cost function - can be error prone - Automation would be nice! - Complex network graph = complex derivative - Dense Layers (Fully connected) - Harder to converge - Number of parameters grows fast! - Overfitting and Parsimony - Learn “well”, generalization capacity - Be efficient in the number of parameters
  • 22. Some Solutions - Calculating the derivative of the Cost function - Software libraries - GPU support for computing vectorial and tensorial data - New Layers Types - Convolution Layers 2D/3D - Dropout layer - Fast activation functions - Faster learning methods - Derived from Stochastic Gradient Descend (SGA) - Weight initializations with Auto-Encoders and RBM
  • 23. Convolutional Networks Idea 1: reuse the weights across while scanning the image Idea 2: subsampling results from layers to layers
  • 24. Fast Activation Functions Idea: don’t use complex exponential functions, linear functions are fast to compute, and easy to differentiate !
  • 25. Dropout Layer, Batch Weight Normalization Dropout: Set randomly some of the input to zero. It improves generalization and makes the network function more robust to errors. Batch Weight Normalization: Normalize the activations of the previous layer at each batch.
  • 26. Efficient Symbolic Differentiation There are good libraries which calculate the derivatives symbolically of an arbitrary number of stacked layers ● efficient symbolic differentiation ● dynamic C code generation ● transparent use of a GPU CNTK
  • 27. Efficient Symbolic Differentiation (2) There are good libraries which calculate the derivatives symbolically of an arbitrary number of stacked layers ● efficient symbolic differentiation ● dynamic C code generation ● transparent use of a GPU >>> import theano >>> import theano.tensor as T >>> from theano import pp >>> x = T.dscalar('x') >>> y = x ** 2 >>> gy = T.grad(y, x) >>> f = theano.function([x], gy) pp(f.maker.fgraph.outputs[0]) '(2.0 * x)'
  • 28. Higher Abstraction Layer: Keras Keras: Deep Learning library for Theano and TensorFlow - Easier to stack layers - Easier to train and test - More ready-made blocks http://keras.io/
  • 29. Example 1: Iris classification Categorize Iris flowers based on - Sepal length/width - Petal length/width 3 classes, Dataset is quite small (150 samples) - Iris Setosa - Iris Versicolour - Iris Virginica input : 4 dimensions output: 3 classes
  • 30. Iris classification: Network model = Sequential() model.add(Dense(15, input_shape=(4,))) model.add(Activation('relu')) model.add(Dropout(0.1)) model.add(Dense(10)) model.add(Activation('relu')) model.add(Dropout(0.1)) model.add(Dense(nb_classes)) model.add(Activation('softmax')) SOFTMAX RELU RELU Setosa Versicolour Virginica Dropout 10% Dropout 10% Train- Test split 80% - 20% Test accuracy: 96%
  • 31. Example 2: telecom customer marketing Semi-synthetic dataset The "churn" data set was developed to predict telecom customer churn based on information about their account. The data files state that the data are "artificial based on claims similar to real world". These data are also contained in the C50 R package. 1 classes (churn) Dataset is quite small (about 3000 samples) 17 input dimensions: State, account length, area code, phone number,international plan,voice mail plan,number vmail messages,total day minutes,total day calls,total day charge,total eve minutes,total eve calls,total eve charge,total night minutes,total night calls,total night charge,total intl minutes,total intl calls,total intl charge,number customer service calls
  • 32. Churn telecom: Network model = Sequential() model.add(Dense(50, input_shape=(17,))) model.add(Activation("hard_sigmoid")) model.add(BatchNormalization()) model.add(Dropout(0.1)) model.add(Dense(10)) model.add(Activation("hard_sigmoid")) model.add(BatchNormalization()) model.add(Dropout(0.1)) model.add(Dense(1)) model.add(Activation(sigmoid)) SOFTMAX RELU RELU Churn No-Churn Dropout 10% Dropout 10% Train- Test split 80% - 20% Test accuracy: 82%
  • 33. Models: Small Data, Big Data - Not all domains have large amount of data - Think of Clinical Tests, or Lengthy/Costly Experimentations - Small specialized data set and Neural Networks - Good for complex non-linear separation of classes Interesting Read: https://medium.com/@ShaliniAnanda1/an-open-letter-to-yann-lecun-22b244fc0a5a#.ngpal1ojx
  • 34. Conclusions - Neural Networks can be used for small data as well - Other methods might be more efficient in this scenario’s - Neural Networks are an extension to GLMs and linear regression - Learn Linear Regression, GLM, SVM as well - Random Forests and Boosted Trees are an alternative - More data = Bigger and better Neural Networks - We have some tools to jump start analysis
  • 35. Connect on Twitter and Linkedin ! Thanks!