Scope for this meetup - Ch3 Problems of the book "Machine Learning: A Probabilistic Perspective" by Kevin Murphy (http://amzn.com/dp/0262018020/)
Sheet containing meta info https://docs.google.com/spreadsheets/d/1INtfzUAeSdCIY0TinoK6em1hzB1O6ubqH16cX3QOcrI/edit#gid=[masked]
Thanks for everyone who turned up for the last meetup!
We're now planning to make this a bi-weekly event. For the next meetup, we will discuss problems from Ch3.
We welcome you to join us, even if you're interested in specific sections of the book. The more people, the more discussions!
If you're interested but can't make it to this event, please get in touch with us on the Boston Data Science Slack http://bit.ly/bostondatascience
Info about the book
This book was recommended to me by my professor at grad school, and it attempts to provide a detailed explanation of the different types of Machine Learning models and algorithms, with a prime focus on Bayesian approach to learning. It doesn’t assume a prior background in statistics, though knowledge of calculus and linear algebra is expected.
This book (and the book club) is mainly intended for people who’ve been in the field for a while, they’ve been using the various Machine Learning models and would love to understand how stuff works under the hood. This knowledge may allow enable to better use or tweak these models to increase performance.
For instance, it can provide answers to questions like -
a) Why is the logistic regression loss called “cross-entropy” function ? Why does it have that equation ?
b) Why is correlation so important ?
Does a 0 correlation mean the variables are independent ? (The answer is no.)
RSVP now and come join us! Also, sign up to the Boston Data Science Slack platform to stay up to date with the book club's schedule and discussions. http://bit.ly/bostondatascience