AI4IMPACT Live Classroom - Advancing with Neural Networks [3 Sessions]

Event Information

Share this event

Date and Time

Refund Policy

Refund Policy

Refunds up to 7 days before event

Eventbrite's fee is nonrefundable.

Event description

Description

We continue to build on the foundation of "Forecasting with Neural Networks". In this course, you will learn and use more advanced techniques for Neural Networks, to help you create better, more accurate forecasts.

To take this class, you should have completed NN101: Forecasting with Neural Networks course.

WHAT YOU WILL LEARN:

  • The concepts of momentum and force inputs
  • Removing noise from your prediction pipeline
  • Ensemble method - bagging and boosting


WHY ATTEND?

  • You want to improve your forecasts
  • You want to solve more challenging real-world prediction problems
  • You want to acquire more advanced skills in your arsenal


LIVE CLASSROOM SCHEDULE:

  • Session 1 - May 14th - 10AM to 11.30AM
  • Session 2 - May 21st - 10AM to 11.30AM
  • Session 3 - May 28th - 10AM to 11.30AM


COURSE OUTLINE


Session 1:

  • Using momentum and force inputs
  • Preprocessing subnetwork for outliers
  • Using a Compression/Decompression subnetwork for noise removal
  • Application to SG Temps ( Part 1 )
  • Summary and Homework


Session 2:

    • Using squared inputs
    • Non-local losses (momentum and force losses)
    • Application to SG Temps ( Part 2 )
    • Application to Energy Demand Forecasting
    • Summary and Homework

Session 3:

  • Ensemble methods - Bagging and Boosting
  • Industry challenge: Predicting Energy Demand (Schneider Electric)



Join our community: http://terra-ai.sg/ai4impact/


TRAINER'S PROFILE:

ARNOLD DORAY is the Head of AI of Terra AI, a predictive technology company. He has 20+ years of experience building global weather forecasting system for mission-critical operations. Arnold has trained 1200 students from NUS, NTU, and SMU in developing predictive AI applications.


Any question? Please contact hi@ai4impact.org.


Date and Time

Location

Online Event

Refund Policy

Refunds up to 7 days before event

Eventbrite's fee is nonrefundable.

Save This Event

Event Saved