Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Machines are Learning: Bringing Powerful Artificial Intelligence Tools to Developers

Machines are Learning: Bringing Powerful Artificial Intelligence Tools to Developers

WAJUG, Liège, November 20th, 2017

Have you always wanted to add predictive capabilities or conversational interfaces to your application, but haven’t been able to find the time or the right technology to get started?
Everybody wants to build smart apps, but only a few are Data Scientists. This session will help you understand machine learning terminology & challenges.

Danilo Poccia

November 20, 2017
Tweet

More Decks by Danilo Poccia

Other Decks in Programming

Transcript

  1. Machines are Learning
    Bringing Powerful Artificial Intelligence to All Developers
    Danilo Poccia
    AWS Technical Evangelist
    @danilop [email protected]
    danilop

    View full-size slide

  2. Credit: Gerry Cranham/Fox Photos/Getty Images
    http://www.telegraph.co.uk/travel/destinations/europe/united-kingdom/england/london/galleries/The-history-of-the-Tube-in-pictures-150-years-of-London-Underground/1939-ticket-examin/

    View full-size slide

  3. 1939 London Underground
    Credit: Gerry Cranham/Fox Photos/Getty Images
    http://www.telegraph.co.uk/travel/destinations/europe/united-kingdom/england/london/galleries/The-history-of-the-Tube-in-pictures-150-years-of-London-Underground/1939-ticket-examin/

    View full-size slide

  4. Data Predictions

    View full-size slide

  5. Data Model Predictions

    View full-size slide

  6. http://www.thehudsonvalley.com/articles/60-years-ago-today-local-technology-demonstrated-artificial-intelligence-for-the-first-time
    1959 Arthur Samuel

    View full-size slide

  7. Machine Learning

    View full-size slide

  8. Machine Learning
    Supervised
    Learning
    Inferring a model
    from labeled
    training data

    View full-size slide

  9. Machine Learning
    Supervised
    Learning
    Unsupervised
    Learning
    Inferring a model
    from labeled
    training data
    Inferring a model
    to describe hidden
    structure from
    unlabeled data

    View full-size slide

  10. Reinforcement
    Learning
    Perform a certain
    goal in a
    dynamic
    environment
    Machine Learning
    Supervised
    Learning
    Unsupervised
    Learning

    View full-size slide

  11. Driving a vehicle
    Playing a game
    against an opponent

    View full-size slide

  12. Tip: Try topic modeling with your own emails ;-)
    Topic Modeling
    Discovering abstract “topics”
    that occur in a collection of documents
    For example, looking for “infrequent” words
    that are used more often in a document

    View full-size slide

  13. Regression “How many bikes will
    be rented tomorrow?”
    Happy, Sad, Angry,
    Confused, Disgusted,
    Surprised, Calm,
    Unknown
    Binary
    Classification
    Multi-Class
    Classification
    “Is this email spam?”
    “What is the
    sentiment of this
    tweet, or of this social
    media comment?”
    1, 0, 100K
    Yes / No
    True / False
    %

    View full-size slide

  14. Training the Model
    Minimizing the Error
    of using the Model on the Labeled Data

    View full-size slide

  15. Validation
    How well is this Model working on New Data?

    View full-size slide

  16. Be Careful of Overfitting

    View full-size slide

  17. Be Careful of Overfitting

    View full-size slide

  18. Be Careful of Overfitting

    View full-size slide

  19. Better Fitting

    View full-size slide

  20. Better Fitting

    View full-size slide

  21. Different Models ⇒ Different Predictions

    View full-size slide

  22. Labeled Data

    View full-size slide

  23. Labeled Data
    70%
    30%
    Training
    Validation

    View full-size slide

  24. Neural
    Networks

    View full-size slide

  25. 1943 Warren McCulloch, Walter Pitts
    Threshold
    Logic
    Units

    View full-size slide

  26. 1962 Frank Rosenblatt
    Perceptron

    View full-size slide


  27. w1
    w2
    w3
    wn
    w0
    =
    output
    weights
    (parameters)
    activation
    function
    input

    View full-size slide

  28. f(∑)
    w1
    w2
    w3
    wn
    w0
    =
    weights
    (parameters)
    activation
    function
    output
    input

    View full-size slide

  29. f(∑)
    input output

    View full-size slide

  30. 1969 Marvin Minsky, Seymour Papert
    Perceptrons:
    An Introduction
    to Computational Geometry
    A perceptron can only solve
    linearly separable functions
    (e.g. no XOR)

    View full-size slide

  31. f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    input
    layer
    hidden
    layer
    output
    layer
    input output
    Multiple Layers
    Lots of Parameters
    Backpropagation

    View full-size slide

  32. Microprocessor Transistor Counts 1971-2011
    Intel Xeon CPU
    28 cores
    NVIDIA V100 GPU
    5,120 CUDA Cores
    640 Tensor Cores
    https://en.wikipedia.org/wiki/Moore's_law

    View full-size slide

  33. LeCun, Gradient-Based
    Learning Applied to Document
    Recognition,1998
    Hinton, A Fast Learning
    Algorithm for Deep Belief
    Nets, 2006
    Bengio, Learning Deep
    Architectures for AI, 2009
    Advances in Research 1998-2009

    View full-size slide

  34. “Stacks of differentiable
    non-linear functions
    with lots of parameters
    solve nearly any predictive
    modeling problem”
    —Jeremy Howard, fast.ai

    View full-size slide

  35. Image
    Processing

    View full-size slide

  36. f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    f(∑)
    output
    How to give images in input
    to a Neural Network?
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  37. Convolution Matrix
    0 0 0
    0 1 0
    0 0 0
    Identity
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  38. Convolution Matrix
    1 0 -1
    2 0 -2
    1 0 -1
    Left Edges
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  39. Convolution Matrix
    -1 0 1
    -2 0 2
    -1 0 1
    Right Edges
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  40. Convolution Matrix
    1 2 1
    0 0 0
    -1 -2 -1
    Top Edges
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  41. Convolution Matrix
    -1 -2 -1
    0 0 0
    1 2 1
    Bottom Edges
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  42. Convolution Matrix
    0.6 -0.6 1.2
    -1.4 1.2 -1.6
    0.8 -1.4 1.6
    Random Values
    Photo by David Iliff. License: CC-BY-SA 3.0
    https://commons.wikimedia.org/wiki/File:Colosseum_in_Rome,_Italy_-_April_2007.jpg

    View full-size slide

  43. Convolutional Neural Networks (CNNs)
    https://en.wikipedia.org/wiki/Convolutional_neural_network

    View full-size slide

  44. ImageNet Classification Error Over Time
    0
    5
    10
    15
    20
    25
    30
    2010 2011 2012 2013 2014 2015 2016 2017
    Classification Error
    CNNs

    View full-size slide

  45. 2012 ImageNet Classification with Deep Convolutional Neural Networks

    View full-size slide

  46. SuperVision: 8 layers, 60M parameters
    0

    View full-size slide

  47. 2013 Visualizing and Understanding Convolutional Networks

    View full-size slide

  48. How Do Neural Networks Learn?
    ?
    More generic and can be reused
    as feature extractor for other visual tasks
    Specific
    to task
    Cat
    Dog
    0

    View full-size slide

  49. The Challenge For Machine Learning: Scale
    Aggressive migration
    New data created on AWS
    PBs of existing data
    Data

    View full-size slide

  50. The Challenge For Machine Learning: Scale
    Tons of GPUs
    Elastic capacity
    Pre-built images
    Aggressive migration
    New data created on AWS
    PBs of existing data
    Data Training

    View full-size slide

  51. The Challenge For Machine Learning: Scale
    Tons of GPUs and CPUs
    Serverless
    At the edge, on IoT Devices
    Tons of GPUs
    Elastic capacity
    Pre-built images
    Aggressive migration
    New data created on AWS
    PBs of existing data
    Data Training Prediction

    View full-size slide

  52. Natural Language Processing Experiment: Topic Modeling
    EC2 Spot Instances
    1.1 Million vCPUs

    View full-size slide

  53. Fulfilment & logistics Search & discovery Existing products New products
    Thousands Of Amazon Engineers Focused On Machine Learning

    View full-size slide

  54. Machine Learning On AWS Today

    View full-size slide

  55. Artificial Intelligence In The Hands Of Every Developer
    S E R V I C E S
    P L A T F O R M S
    E N G I N E S
    I N F R A S T R U C T U R E
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow

    View full-size slide

  56. Early Detection
    Of Diabetic
    Complications

    View full-size slide

  57. FDA-approved
    Medical Imaging

    View full-size slide

  58. Sports Analytics

    View full-size slide

  59. Autonomous Driving Systems

    View full-size slide

  60. Real Time, Per Pixel Object Segmentation

    View full-size slide

  61. Centimeter-accurate Positioning

    View full-size slide

  62. Computation Knowledge Engine

    View full-size slide

  63. S E R V I C E S
    P L A T F O R M S
    E N G I N E S
    I N F R A S T R U C T U R E
    Amazon ML Spark & EMR Kinesis Batch ECS
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow
    Artificial Intelligence In The Hands Of Every Developer

    View full-size slide

  64. S E R V I C E S
    P L A T F O R M S
    Vision
    Amazon Rekognition
    E N G I N E S
    I N F R A S T R U C T U R E
    Amazon ML Spark & EMR Kinesis Batch ECS
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow
    Artificial Intelligence In The Hands Of Every Developer

    View full-size slide

  65. Mona Lisa
    (Leonardo da Vinci)

    View full-size slide

  66. Mona Lisa
    (Prado's version)

    View full-size slide

  67. Portrait of
    Maddalena Doni
    (Raphael)

    View full-size slide

  68. Bynder allows you to easily create, find and use content
    for branding automation and marketing solutions.
    With our new AI capabilities,
    Bynder’s software… now allows
    users to save hours of admin
    labor when uploading and
    organizing their files, adding
    exponentially more value.
    Chris Hall
    CEO, Bynder


    With Rekognition, Bynder revolutionizes marketing admin tasks with AI capabilities

    View full-size slide

  69. S E R V I C E S
    P L A T F O R M S
    Speech
    Amazon Polly
    Vision
    Amazon Rekognition
    E N G I N E S
    I N F R A S T R U C T U R E
    Amazon ML Spark & EMR Kinesis Batch ECS
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow
    Artificial Intelligence In The Hands Of Every Developer

    View full-size slide

  70. Generate Lifelike Speech With Amazon Polly
    25 languages
    “The temperature in
    Milanis 16 degrees
    Celsius”
    “The temperature
    in Milan is 16˚C”
    Amazon
    Polly
    52 voices

    View full-size slide

  71. aws polly synthesize-speech
    --text "It was nice to live such a wonderful live show."
    --output-format mp3
    --voice-id Joanna
    --text-type text
    output.mp3

    View full-size slide

  72. Duolingo voices its language learning service Using Polly
    Duolingo is a free language learning service where users
    help translate the web and rate translations.
    With Amazon Polly our users
    benefit from the most lifelike
    Text-to-Speech voices
    available on the market.
    Severin Hacker
    CTO, Duolingo

    “ • Spoken language crucial for
    language learning
    • Accurate pronunciation matters
    • Faster iteration thanks to TTS
    • As good as natural human speech

    View full-size slide



  73. Royal National Institute of Blind People creates and
    distributes accessible information in the form of
    synthesized content
    Amazon Polly delivers
    incredibly lifelike voices which
    captivate and engage our
    readers.
    John Worsfold
    Solutions Implementation Manager, RNIB
    • RNIB delivers largest library of
    audiobooks in the UK for nearly 2 million
    people with sight loss
    • Naturalness of generated speech is
    critical to captivate and engage readers
    • No restrictions on speech redistributions
    enables RNIB to create and distribute
    accessible information in a form of
    synthesized content
    RNIB provides the largest library in the UK for people with sight loss

    View full-size slide

  74. S E R V I C E S
    P L A T F O R M S
    Chat
    Amazon Lex
    Speech
    Amazon Polly
    Vision
    Amazon Rekognition
    E N G I N E S
    I N F R A S T R U C T U R E
    Amazon ML Spark & EMR Kinesis Batch ECS
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow
    Artificial Intelligence In The Hands Of Every Developer

    View full-size slide

  75. Amazon Lex
    Speech recognition and natural language understanding
    Automatic speech recognition
    Natural language understanding
    “What’s the weather
    forecast?”
    Weather
    forecast
    Amazon Lex

    View full-size slide

  76. Amazon Lex
    Speech recognition and natural language understanding
    “It will be
    sunny
    and 16C”
    Automatic speech recognition
    Natural language understanding
    “What’s the weather
    forecast?”
    Weather
    forecast
    Amazon Lex

    View full-size slide

  77. “It will be sunny
    and 16 degrees
    Celsius”
    Amazon Polly
    Amazon Lex
    “It will be
    sunny
    and 16C”
    Automatic speech recognition
    Natural language understanding
    “What’s the weather
    forecast?”
    Weather
    forecast
    Speech recognition and natural language understanding
    Amazon Lex

    View full-size slide



  78. Finding missing persons:
    ~100,000 active missing
    persons cases in the U.S.
    at any given time
    ~60% are adults,
    ~40% are children
    • Motorola Solutions applies Amazon
    Rekognition, Amazon Polly and Amazon
    Lex
    • Image analytics and facial recognition
    can continually monitor for missing
    persons
    • Tools that understand natural language
    can enable officers to keep eyes up and
    hands free
    Motorola Solutions is using AI to help finding missing persons
    Motorola Solutions keeps utility workers connected and
    visible to each other with real-time voice and data
    communication across the smart grid.

    View full-size slide

  79. S E R V I C E S
    P L A T F O R M S
    Chat
    Amazon Lex
    Speech
    Amazon Polly
    Vision
    Amazon Rekognition
    E N G I N E S
    I N F R A S T R U C T U R E
    Amazon ML Spark & EMR Kinesis Batch ECS
    GPU CPU IoT Mobile
    Apache MXNet Caffe 2 Theano PyTorch CNTK
    TensorFlow

    View full-size slide

  80. There’s Never Been A Better Time To Build Smart Apps

    View full-size slide

  81. https://github.com/danilop/security-camera

    View full-size slide

  82. Machines are Learning
    Bringing Powerful Artificial Intelligence to All Developers
    Danilo Poccia
    AWS Technical Evangelist
    @danilop [email protected]
    danilop

    View full-size slide