Machine Learning

Clinical Informatics

Goal: 
Extract clinically useful medical knowledge from large amounts of medical data stored in Electronic Health Records (EHR) toward a better diagnosis and prevention of diseases and conditions.
Clinical Informatics Group (CING) at CCLS

Our group is dedicated to developing machine learning algorithms to leverage large medical data stored in Electronic Health Records. Our approach brings to bear new methods to derive accurate, multi-dimensional models from large collections of observational data.

Our research team brings together machine learning, natural language processing experts, database architects and programmers from the Center for Computational Learning Systems (CCLS) at Columbia University along with clinicians from Columbia University Medical Center (CUMC).

Project link:

http://www1.ccls.columbia.edu/~ansaf/CING/

Climate Informatics

Goal: 
Forge collaborations between machine learning and climate science, in order to accelerate progress in answering pressing questions in climate science.
Collaborations between machine learning and climate science, in order to accelerate progress in answering pressing questions in climate science

The threat of climate change is one of the greatest challenges currently facing society.  Given the profound impact machine learning has made on the natural sciences to which it has been applied,

Project Twiki link: 
https://power.ldeo.columbia.edu/twiki/bin/view/Main/ClimateInformatics

An Advanced Learning Paradigm: Learning Using Hidden Information

Goal: 
Develop algorithms in the SVM family that allow extra information to be used effectively during training, with the understanding that this extra information will not be available during actual operation
Learning extra information like structural homologies between proteins in a system designed to predict structure from amino acid sequences

Estimation of Mean Time Between Failures (MTBF) of Electrical Feeders and Related Components

The project aims to estimate the time between (and to) failures of primary distribution feeders and their components (such as sections and joints).

In the New York City Power Grid, electricity is transmitted via primary distribution feeders between the high voltage transmission system and the household-voltage secondary system. These feeders are susceptible to different kinds of failures such as emergency isolation caused by automatic substation relays (Open Autos), failing on test, maintainence crew noticing problems and scheduled work on different sections of the feeder.

EEGMine: A Distributed Framework for Learning on iEEG Data

Goal: 
This project aims to develop a distributed framework for Data Management and Machine Learning on iEEG data obtained from Epilepsy patients
Distributed Data Mining (DDM) on iEEG data

Sample of EEG Graph from a patient

Project Twiki link: 
https://power.ldeo.columbia.edu/twiki/bin/view/SeizurePrediction/EEGMineDocs

Rigorous Learning

Rigorous Learning

Rigorous Learning
A National Science Foundation Grant (CCR-0325463):
New Directives in Predictive Learning: Rigourous Learning Machines.
Senior Team Members: Vladimir Vapnik, David Waltz, David Hessing, Eugine Ie, Alex Gammerman, Nataliya Pavlovich

Concept Learning for Vision and Planning

Concept Learning for Vision and Planning

Concept Learning for Vision and Planning
A National Science Foundation SGER(Small Grant for Exploratory Research): Computational Basis for Cognition: Concept Learning for Vision and Planning grant to study the Computational Basis for Cognition.
Senior Team Members: David Waltz, Yoav Freund, Eric Baum

Syndicate content