@SubinAlex9@johnmwinn@tommy_da_cat We did aim to be suitable for beginners and introduce all concepts needed - have a read of this introductory chapter and see for yourself: https://t.co/8INya89Caq
My new book Model-Based Machine Learning with Chris Bishop, Tom Diethe @tommy_da_cat, John Guiver and Yordan Zaykov is now available.
Huge thanks to everyone who helped put the book together!
Order from https://t.co/QrtGyuYfHQ
#MachineLearning#newbook
So many readers of the online book have given feedback, corrected typos and generally been hugely encouraging and positive. I've tried to thank as many of you as possible in the acknowledgements.
I hope you will see this tweet and know that your input made a huge difference!
Statistical models are mathematical representations of the real world. So if you lose sight of either the real world *or what the math you're applying is actually doing*, your model will be prone to failure.
At long last, the final chapter of the book is here - 'How to Read a Model' - diving into the assumptions, strengths and weaknesses of four common ML models.
Please take a look and send feedback! https://t.co/FGjHbuNOCY
A few times a year, I get asked to be a judge of student statistical projects in politics or sports. While the students are very bright, they spend WAY too much time using fancy statistical methods and not enough time framing the right questions and contextualizing their answers.
"Model-based vs. model-free" is possibly the worst red herring in all of machine learning. *All* of RL is model based due to assuming an MDP structure and using the Bellman equations. 1/3
@ESA_EO Other brilliant introductory material for the general public (about the "valley of right enough" and the probabilistic view of ML) is a book written by @ChrisBishopMSFT @johnmwinn and others, with the first chapter titled "A Murder Mystery"
https://t.co/dv5yQKCehC
In our final chapter, “How to read a Model”, we’ll look at existing ML models to see what assumptions they make, when they can be applied and how they can be extended. We’re looking for suggestions of models to look at - if you have ideas, please tweet them to us!!
Delighted to announce that I'm now leading the "All Data AI (ADA)" theme at Microsoft Research Cambridge.
If you love turning ideas into reality, check out our jobs postings https://t.co/plEJOY8tTn