Data Mining Reading Group
What?: Data Mining Reading Group is a bunch of data mining enthusiasts where we will present a paper to the rest of us every week. Each one of us should try to read this paper critically. The presenter should spend more time thinking about how he/she would like to present the material.
Why?: Several Reasons
- Learn about the state-of-the-art research papers
- Improve on reading papers.
- Improve on presentation skills.
When ? Every Fridays at 11:00 - 12:00 pm.
When ? Large CS Conference Room/Watch the Email
Please edit this table and enter a date when you can present a paper.
Faculty Members (Alphabetical Order):
- Daniel Barbara
- Carlotta Domeniconi
- Jessica Lin
- Jana Kosecka
- Huzefa Rangwala
Theme for the Semester: Structured Output Learning/Mutli-Task/Multi-instance learning
|09.16.2011:||Sam Blasiak||MEDLDA by Zhu et. al.|
|09.30.2011:||Jessica Lin||Mining Massive Time Series Database.|
|10.14.2011 (11-12):||Guoxian Yu||Semi-Supervised Learning with High Dimensional Data|
|09.10.2010||Sam B||Latent Dirichlet Allocation by David Blei, Andrew Ng and Michael Jordan Sam will be presenting a detailed derivation of the LDA variational algorithm|
|09.30.2010 [THU] #4201/11-12 noon||Pu W||Introduction to Monte Carlo Methods by Mackay|
|10.08.2010 [FRI] #4201||Muzzamil S||Characterizing Microblogs with Topic Models by Ramage et. al.|
|10.25.2010||Syed F||K-Means on Commodity GPUs with CUDA by Hong-Tao et. al.|
Leveraging Sequence Classication by Taxonomy-based Multitask Learning
by Widmer et. al
|11.05.2010||Tanwishta S||A kernel method for unsupervised structured network inference by Lippert et. al|
|11.12.2010||Zeehasham R||Extreme Learning Machine: A New Learning Scheme of Feedforward Neural Networks by Huang et. al|
|11.19.2010||Chaitanya Y||03.26.2010, Room #4801||Anveshi C||Classification of protein sequences by means of irredundant patterns by Comin et. al.|
|04.02.2010||Sam B||Conditional random fields: Probabilistic models for segmenting and labeling sequence data by Lafferty|
|04.09.2010||Zeehasham R||Using Highly Expressive Contrast Patterns for Classification - Is It Worthwhile ?|
Potential Papers that can be presented (Please feel free to pick one you had like not on this list):
- J. Chang and D. Blei. Relational Topic Models for Document Networks . Artificial Intelligence and Statistics, 2009.
- Large Margin Semi-supervised Learning by J Wang
- Conditional random fields: Probabilistic models for segmenting and labeling sequence data by Lafferty
- An ensemble framework for clustering protein–protein interaction networks by Asur et. al
- Kernel methods for predicting protein–protein interactions by Ben-Hur et. al
- Using Highly Expressive Contrast Patterns for Classification - Is It Worthwhile ? Elsa Loekito and James Bailey. Proceedings of the 13th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD-09),
- Mining minimal distinguishing subsequence patterns with gap constraints
by Xiaonan Ji, James Bailey & Guozhu Dong
- Fast subtree kernels on graphs by N. Shervashidze, K. Borgwardt