Please log in to watch this conference skillscast.
Ensemble methods are a great way to improve the accuracy of deep learning models. However, current ensemble methods all treat the underlying base learner as a "black box".
Join Alan as he illustrates a new type of ensemble method which makes use of information about how the deep learning algorithm is structured, in order to improve training times, diversity generation, and ultimately accuracy.
Alan will also give examples of existing methods from his research at Birkbeck, and provide experimental results.
YOU MAY ALSO LIKE:
- Fast, scalable Deep Learning Ensembles (SkillsCast recorded in November 2016)
- Leonardo De Marchi's Deep Learning Fundamentals (in London on 22nd - 23rd October 2019)
- Brian Sletten's Data Science with Python Workshop (in London on 18th - 20th November 2019)
- Scala eXchange London 2019 (in London on 12th - 13th December 2019)
- Practical ML 2020 (in London on 2nd - 3rd July 2020)
- PowerBI The Must-Know Speed Up Tricks (in London on 17th October 2019)
- Countdown to Big Data LDN (in London on 17th October 2019)
- Porcupine: Flows Your Rows with Arrows (SkillsCast recorded in October 2019)
- Automating Elaborate-Transform-Load for Busy Data Scientists (SkillsCast recorded in October 2019)
White-box Deep Learning Ensembles - Advanced
Alan Mosca is a part-time research student at Birkbeck, University of London. He also holds a full time position as Senior Data Engineer at Sendence. His research focus is on Deep Learning Ensembles and improvements to optimisation algorithms in Deep Learning. He was previously at Wadhwani Asset Management, Jane Street Capital, and several software companies.