Gaussian Processes for Big Data problems - Intermediate

6th July 2017 in London at CodeNode

There are 43 other SkillsCasts available from Infiniteconf 2017 - the conference on Big Data and Fast Data

Please log in to watch this conference skillscast.

643587683 640

Gaussian Processes (GPs) underpin a range of algorithms for regression, classification and unsupervised learning. GPs are mathematically equivalent to many well known models, including Bayesian linear models, spline models, large neural networks (under suitable conditions), and are closely related to others, such as support vector machines. A rich theory also exists for GP models in the time series analysis literature.

Seen the flexibility and the advantages of GP models, it is not surprising that recent research has been focused on extending GP models to BigData problems.

Join Roberta as she shares an overview of GPs in a machine learning context and how they can be used for BigData problems. Explore the basic definition of a Gaussian distribution and GPs, with particular attention to covariance functions. Roberta will then explain how GPs are used for inference problems and present relevant application areas, their advantages and their limitations.


Thanks to our sponsors

Gaussian Processes for Big Data problems - Intermediate

Roberta Cretella

I have a Master in mathematics from "La Sapienza, Università di Roma.