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

AI: The next frontier by Amparo Alonso at Big Data Spain 2017

AI: The next frontier by Amparo Alonso at Big Data Spain 2017

AI: The next frontier

https://www.bigdataspain.org/2017/talk/ai-next-frontier

Big Data Spain 2017
November 16th - 17th Kinépolis Madrid

Big Data Spain

December 05, 2017
Tweet

More Decks by Big Data Spain

Other Decks in Technology

Transcript

  1. A.I.: The next frontier
    Amparo Alonso Betanzos
    CITIC-UDC
    Grupo LIDIA

    View full-size slide

  2. The Primeval Soup: The perfect storm

    View full-size slide

  3. Batch
    Streaming
    Aman Naimat.
    “The new Artificial Intelligence Market.
    The Big data Market”. O´Reilly, 2016
     During 2017 the tendency of data
    generation has showed sustained growth.
     The appetite of corporates, industry and
    public sector for data driven initiatives has
    not decreased.
     There is a change of landscape that by
    2017 has started to become apparent.

    View full-size slide

  4. Data Industry Landscape

    View full-size slide

  5. Infrastructure Challenges
    Data storage
    High
    performance
    in
    interchange
    and sharing
    Data format
    and protocols

    View full-size slide

  6. Advancing hardware

    View full-size slide

  7. Regulation and Ethics

    View full-size slide

  8. Data rich
    vs Data
    poor
    Confidentialit
    y and
    scientific
    transparency
    Reproducibilit
    y Free data
    https://www.linkedin.com/pulse/national-artificial-intelligence-research-development-nco-
    nitrd/

    View full-size slide

  9. High dimensionality data
    Sparse data
    Heterogeneous data
    Missing data
    Noisy data
    Adversarial data
    Untrustworthy data
    Data Science

    View full-size slide

  10. • Machine Learning is as valuable as how exploitable its results
    are.
    • Lagging behind in some areas:
    • Visualization of clusters
    • Data drift
    • Results Assurance
    • Biased data
    2017 Big Data Coruña. Statistical inference for big-but-biased data
    https://www.youtube.com/watch?v=luTJbX3aVKA
    More work
    is needed
    on:
    • Feature engineering
    • Regression
    • Anomaly detection
    • Practical non convex optimization
    • Effective parameter selection
    • Scalable transfer learning
    • Data integration
    • Data visualization
    Reliable Machine Learning

    View full-size slide

  11. Feature
    Engineering
    Distributed FS algorithms
    Missing Data
    Heterogeneou
    s data
    Unbalanced
    data

    View full-size slide

  12. Normalized Discounted Cumulative Gain (NDC

    View full-size slide

  13. • MNIST, 256 relevant features(576pixels)
    • 20% missing (MAR)
    • Imputation using median and SVD (Singular Value Decomposition)
    B. Seijo-Pardo, A. Alonso-Betanzos, K. Bennett, V. Bolón-Canedo, I. Guyon, M. Saeed. Analysis of imputation
    bias for feature selection with missing data. ESANN 2018

    View full-size slide

  14. FS Original
    FS Median Imputation
    FS, SVD imputation

    View full-size slide

  15. Size matters

    View full-size slide

  16. • The study of methodologies that increase the
    scalability of ML principles and algorithms.
    • Scalability should be seen as an abstract concept
    that not only includes the case of dealing with
    huge amounts of data points.
    • Just measuring the challenge in storage units will
    be a narrow minded view that will be oblivious to
    the challenge that current times is putting on the
    shoulders of ML
    Networks of AI systems
    Scalability

    View full-size slide

  17. • Models that can learn under
    privacy and anonimity
    constraints
    • Share parameter values, not
    data
    • Using aggregated data
    • Adequate accuracy?
    • Private data reconstruction?
    Privacy-preserving
    ML
    D. Fernández-Francos, O. Fontenla-Romero, A. Alonso-Betanzos. One-class
    convex hull-based algorithm for classification in distributed environments.
    IEEE Transactions on Systems, Man and Cybernetics: Systems (in press)

    View full-size slide

  18. Learning to Learn
    http://bair.berkeley.edu/blog/2017/07/18/learning-to-learn/

    View full-size slide

  19. https://spectrum.ieee.org/static/ai-vs-doctors
    Narrow niche vs General

    View full-size slide

  20. “Armed with machine learning, a manager becomes a supermanager, a scientist a superscie
    a superengineer. The future belongs to those who understand at a very deep level how to c
    expertise with what algorithms do best.”
    Pedro Domingos, The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake
    https://www.itnonline.com/content/
    new-report-highlights-five-reasons-why-radiology-needs-artificial-intelligence
    Human-in-the-loop

    View full-size slide

  21. • Deep Learning is not the AI future, https://
    www.kdnuggets.com/2017/08/deep-learning-not-ai-futu
    re.html
    • The National AI R&D Strategic plan (USA)
    https
    ://www.linkedin.com/pulse/national-artificial-intelligenc
    e-research-development-nco-nitrd
    /
    • General Data Protection Regulation, UE
    http://ec.europa.eu/justice/data-
    protection/reform/files/regulation_oj_en.pdf
    Explainabilit
    y

    View full-size slide

  22. Transportation
    service robots
    Public safety, security
    AI Applications

    View full-size slide

  23. Education
    Low-resource communities
    AI Applications

    View full-size slide

  24. Entertainment
    Social risk of diminishing interpersonal interacti

    View full-size slide

  25. pplications: Employment and workp

    View full-size slide

  26. The 6 Laws proposed by EU
    All intelligent machine should have an
    emergency switch
    An intelligent machine could not
    damage a human being
    It is forbidden to establish emotional
    links with a machine or electronic
    person
    The biggest machines should have an
    obligatory insurance
    Electronic persons will have rights
    and obligations.
    Electronic persons and machines
    should pay taxes
    http://www.europarl.europa.eu/news/es/news-
    room/20170109STO57505/delvaux-propone-normas-
    europeas-para-la-rob%C3%B3tica-y-un-seguro-obligatorio
    http://computerhoy.com/noticias/life/e
    stas-son-seis-leyes-robotica-que-
    propone-ue-56972

    View full-size slide

  27. 6,3% (16% in Software Industry)

    View full-size slide

  28. A.I.: The next frontier
    Amparo Alonso Betanzos
    CITIC-UDC
    Grupo LIDIA

    View full-size slide