Global AI and Data Science

Global AI & Data Science

Train, tune and distribute models with generative AI and machine learning capabilities

 View Only
Back to discussions
Expand all | Collapse all

Machine Learning: Lessons Learned from the Enterprise: Chat with The Lab Webinar Series

  • 1.  Machine Learning: Lessons Learned from the Enterprise: Chat with The Lab Webinar Series

    Posted Wed May 06, 2020 10:24 PM
    Edited by System Admin Fri January 20, 2023 04:37 PM
    There's a huge difference between the purely academic exercise of training Machine Learning models versus building end-to-end Data Science solutions to real enterprise problems.

    This presentation summarizes the lessons learned after three years of our team engaging with dozens of enterprise clients from different industries including manufacturing, financial services, retail, entertainment, healthcare, etc.

    • What are the most common ML problems faced by the enterprise?
    • What is beyond training an ML model?
    • How do you address data preparation?
    • How do you scale to large datasets?
    • Why is feature engineering so crucial?
    • How do you go from a model to a fully capable system in production?
    • Do you need a Data Science platform if every single data science tool is available in the open source?
    These are some of the questions that will be addressed, exposing some challenges, pitfalls, and best practices through specific industry examples.

    On Demand Recording: here 
    ------------------------------
    JORGE CASTANON
    Chat with labs webinar series: https://ibm.co/Chat-With-The-Lab-Webinar
    ------------------------------
    #GlobalAIandDataScience
    #GlobalDataScience


  • 2.  RE: Machine Learning: Lessons Learned from the Enterprise: Chat with The Lab Webinar Series

    Posted Thu May 14, 2020 01:39 PM

    When "analytics" first started out (quite some time ago,  I'm thinking of IBM Intelligent Data Miner and Text Miner products),  it was far more common to make transparent what algorithms were being used and to surface some of their inherent risks and biases.  This helped remind the user to try other models and methods.  Also  the many feeds-and-speeds inherent in these were more apparent and could be tuned.

    We seem to have gone to much more a "black box" approach not surfacing what's being done nor the potential risks/drawbacks of particular algorithms (and techniques) in certain circumstances.

    Thoughts as to why this happened?



    ------------------------------
    Kevin Minerley
    Sr Software Engineer (RAS and Resiliency)
    IBM
    POUGHKEEPSIE NY
    845-435-7430
    ------------------------------

    Original Message


  • 3.  RE: Machine Learning: Lessons Learned from the Enterprise: Chat with The Lab Webinar Series

    Posted Thu May 14, 2020 02:57 PM
    Hello, 
    You can find the resources mentioned during the presentation below and you can watch On Demand Recording: here 

    Additional Resources

    ML contests and resources:
    www.kaggle.com

    Deep learning intuition:
    https://www.youtube.com/watch?v=aircAruvnKk&t=1s

    Deep learning course:
    https://www.coursera.org/specializations/deep-learning

    Please post any of your questions below! 

    Thanks, 


    ------------------------------
    JORGE CASTANON
    Chat with labs webinar series: https://ibm.co/Chat-With-The-Lab-Webinar
    ------------------------------

    Original Message


Related Content