DeepAI
Log In Sign Up

Notes on Icebreaker

04/26/2020
by   Shalin Shah, et al.
0

Icebreaker [1] is new research from MSR that is able to achieve state of the art performance on inference in which there is inherent missing data. Using mutual information, Icebreaker is able to suggest which values in the data to impute for maximum benefit. These notes are an amalgamation of information from various articles and tutorials including autoencoders, variational inference, variational autoencoders, the evidence lower bound, set based learning and finally leading to Icebreaker. References are provided whenever appropriate. There may be factual errors and typos in these notes. Please send them to the author.

READ FULL TEXT

page 1

page 2

page 3

page 4

10/05/2018

Doubly Semi-Implicit Variational Inference

We extend the existing framework of semi-implicit variational inference ...
06/28/2019

The Thermodynamic Variational Objective

We introduce the thermodynamic variational objective (TVO) for learning ...
12/01/2021

Lecture notes on complexity of quantifier elimination over the reals

These are lecture notes for a course I gave in mid-1990s for MSc student...
12/01/2020

Mutual Information Constraints for Monte-Carlo Objectives

A common failure mode of density models trained as variational autoencod...
09/29/2015

Variational Information Maximisation for Intrinsically Motivated Reinforcement Learning

The mutual information is a core statistical quantity that has applicati...
11/17/2020

Recursive Inference for Variational Autoencoders

Inference networks of traditional Variational Autoencoders (VAEs) are ty...
01/15/2021

Efficient Semi-Implicit Variational Inference

In this paper, we propose CI-VI an efficient and scalable solver for sem...