Empirical Inference Journal Club (EIJC)

What is EIJC?

Empirical Inference Journal Club (EIJC) aims to provide a platform for PhD students, post-doctoral fellows, and members of the Department of Empirical Inference, Max Planck Institute for Intelligent Systems to discuss and exchange ideas on the cutting-edge researches involving the empirical inference. The general topics of interest include, but not limited to

The topics of interest are subjected to change proposed by the EIJC participants.

Structure of EIJC

The ultimate goal of the EIJC to make sure that each participant comes with something, and then goes back with something more. The structure of the EIJC is very simple and consisting of three steps:

Therefore, the contributions of each of the participants are very vital to the success of the journal club.


The initial plan is to have a two-hour meeting on a weekly basis. We meet every Monday from 2pm to 4pm at the AGBS seminar room.

Aug 1, 201211am-1pmA survey of statistical network models Introduction and static network models
Aug 2, 20123.30pm-4.30pmA survey of statistical network modelscontinue
Aug 8, 201211am-1pm A survey of statistical network models dynamic network modelling
Aug 29, 201211am-1pmA state-space mixed membership blockmodel for dynamic network tomography introduction to the model
Sep 12, 201211am-1pmA state-space mixed membership blockmodel for dynamic network tomographyvariational inference
Sep 26, 201211am-1pmMachine learning that matters
Oct 3, 201211am-1pmExogeneity
Jan 14, 201311am-1pmIntroduction to Deep Learning and Its Historya short introduction to perceptron by Krikamol
Jan 21, 201311am-1pmRestricted Boltzmann Machine and Constrastive Divergence

[paper 1] [paper 2]

an informal presentation of artificial neural network and backpropagation by Christopher
Jan 28, 20132pm-4pmStacked Denoising Autoencoders

[paper 1] [paper 2]

Feb 4, 20132pm-4pmRepresentation Learning and Some Insights on Deep Learning

[paper 1][paper 2][paper 3][paper 4]

Feb 11, 20132pm-4pmDeep Learning through Sparsity and Energy-based Models

[paper 1] [paper 2]

a brief introduction to convolutional nets by Chris
Feb 18, 20132pm-4pm
Semi-supervised Embedding, Sparse Coding, and Object Recognition

[paper 1, paper 2, paper 3]

Feb 25, 20132pm-4pm
Large-Scale Unsupervised Learning

[paper 1, paper 2]

Mar 4, 20132pm-4pm
Learning the Structure of Deep Sparse Graphical Models


Mar 11, 20132pm-4pm
Deep Gaussian Processes


A special session with Neil Lawrence
Mar 18, 20132pm-4pm
Kernel Methods for Deep Learning



The previous and current participants of the EIJC include

If you are interested in joining the EIJC, please write an email to krikamol@tuebingen.mpg.de