Feb 28, 2010 NEMO data meta-analysis: Application of NEMO analysis workflow to consortium datasets (redux)

Slides:



Advertisements
Similar presentations
Data Mining and the Web Susan Dumais Microsoft Research KDD97 Panel - Aug 17, 1997.
Advertisements

NEMO ERP Analysis Toolkit ERP Metric Extraction An Overview.
Principal Component Analysis Based on L1-Norm Maximization Nojun Kwak IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
Haishan Liu.  to identify correspondences (“mappings”) between ERP patterns derived from:  different data decomposition methods (temporal PCA and spatial.
February 11, 2011 Overview of All-Hands Meeting Agenda Gwen Frishkoff
Jianwei Lu1 Information Extraction from Event Announcements Student: Jianwei Lu ( ) Supervisor: Robert Dale.
IVITA Workshop Summary Session 1: interactive text analytics (Session chair: Professor Huamin Qu) a) HARVEST: An Intelligent Visual Analytic Tool for the.
MAPPING RESULTS Experiments were performed on two simulated datasets, each using both metric sets. Cross spatial join: calculate the Euclidean distance.
OHBM Morning Workshop – June 20, 2009 Neurocognitive ontologies: Methods for sharing and integration of human brain data Neural ElectroMagnetic Ontologies.
Haishan Liu 1, Gwen Frishkoff 2, Robert Frank 1, Dejing Dou 1 1 University of Oregon 2 Georgia State University.
February 11, 2011 Overview of All-Hands Meeting Agenda Gwen Frishkoff
NEMO ERP Analysis Toolkit ERP Pattern Segmentation An Overview.
Haishan Liu 1, Gwen Frishkoff 2, Robert Frank 1, Dejing Dou 1 1 University of Oregon 2 Georgia State University.
September 16, 2009 NEMO OWL ontologies: Viewing & editing OWL/RDF files, part II
NEMO Data Analysis Workflow and MATLAB Tools
Rubber Hits the Road: Why NEMO needs RDF Paea LePendu Stanford Center for Biomedical Informatics Research National Center for Biomedical Ontology (NCBO)
August 20, 2009 NEMO Year 1: From Theory to Application — Ontology-based analysis of ERP data
IR & Metadata. Metadata Didn’t we already talk about this? We discussed what metadata is and its types –Data about data –Descriptive metadata is external.
Report on Intrusion Detection and Data Fusion By Ganesh Godavari.
Chapter 6 Database Design
The Data Mining Visual Environment Motivation Major problems with existing DM systems They are based on non-extensible frameworks. They provide a non-uniform.
Advanced Topics COMP163: Database Management Systems University of the Pacific December 9, 2008.
February 26, 2010 NEMO All-Hands Meeting: Overview of Day 1
Presented by Zeehasham Rasheed
Ranking by Odds Ratio A Probability Model Approach let be a Boolean random variable: document d is relevant to query q otherwise Consider document d as.
Latent Semantic Analysis (LSA). Introduction to LSA Learning Model Uses Singular Value Decomposition (SVD) to simulate human learning of word and passage.
6 Chapter 6 Database Design Hachim Haddouti. 6 2 Hachim Haddouti and Rob & Coronel, Ch6 In this chapter, you will learn: That successful database design.
The Neural ElectroMagnetic Ontology (NEMO) System: Design & Implementation of a Sharable EEG/MEG Database with ERP ontologies G. A. Frishkoff 1,3 D. Dou.
Tomer Sagi and Avigdor Gal Technion - Israel Institute of Technology Non-binary Evaluation for Schema Matching ER 2012 October 2012, Florence.
MDC Open Information Model West Virginia University CS486 Presentation Feb 18, 2000 Lijian Liu (OIM:
AdvisorStudent Dr. Jia Li Shaojun Liu Dept. of Computer Science and Engineering, Oakland University 3D Shape Classification Using Conformal Mapping In.
Database System Development Lifecycle © Pearson Education Limited 1995, 2005.
ERP DATA ACQUISITION & PREPROCESSING EEG Acquisition: 256 scalp sites; vertex recording reference (Geodesic Sensor Net)..01 Hz to 100 Hz analogue filter;
July 17, 2009 NEMO Year 1: Overview & Planning
1 Dejing Dou Computer and Information Science University of Oregon, Eugene, Oregon September, Kent State University.
Mining Discriminative Components With Low-Rank and Sparsity Constraints for Face Recognition Qiang Zhang, Baoxin Li Computer Science and Engineering Arizona.
Week 4 Lecture Part 3 of 3 Database Design Samuel ConnSamuel Conn, Faculty Suggestions for using the Lecture Slides.
Scott Duvall, Brett South, Stéphane Meystre A Hands-on Introduction to Natural Language Processing in Healthcare Annotation as a Central Task for Development.
Mining Shifting-and-Scaling Co-Regulation Patterns on Gene Expression Profiles Jin Chen Sep 2012.
1 Technologies for (semi-) automatic metadata creation Diana Maynard.
Intelligent Database Systems Lab Presenter : WU, MIN-CONG Authors : Jorge Villalon and Rafael A. Calvo 2011, EST Concept Maps as Cognitive Visualizations.
Report on Intrusion Detection and Data Fusion By Ganesh Godavari.
2 nd International Conference on Biomedical Ontology (ICBO’11) Ontology-Based Analysis of Event-Related Potentials Gwen Frishkoff 12, Robert Frank 2, Paea.
1/26/2004TCSS545A Isabelle Bichindaritz1 Database Management Systems Design Methodology.
6.1 © 2010 by Prentice Hall 6 Chapter Foundations of Business Intelligence: Databases and Information Management.
Mingyang Zhu, Huaijiang Sun, Zhigang Deng Quaternion Space Sparse Decomposition for Motion Compression and Retrieval SCA 2012.
EXPERIMENT DESIGN  Variations in Channel Density  The original 256-channel data were downsampled:  127 channel datasets  69 channels datasets  34.
Region-Based Saliency Detection and Its Application in Object Recognition IEEE TRANSACTIONS ON CIRCUITS AND SYSTEM FOR VIDEO TECHNOLOGY, VOL. 24 NO. 5,
1 Masters Thesis Presentation By Debotosh Dey AUTOMATIC CONSTRUCTION OF HASHTAGS HIERARCHIES UNIVERSITAT ROVIRA I VIRGILI Tarragona, June 2015 Supervised.
Rubber Hits the Road: How RDF benefits NEMO
Introduction to Data Mining by Yen-Hsien Lee Department of Information Management College of Management National Sun Yat-Sen University March 4, 2003.
EEG DATA EEG Acquisition: 256 scalp sites; vertex recording reference (Geodesic Sensor Net)..01 Hz to 100 Hz analogue filter; 250 samples/sec. EEG Preprocessing:
Achieving Semantic Interoperability at the World Bank Designing the Information Architecture and Programmatically Processing Information Denise Bedford.
Towards Unifying Vector and Raster Data Models for Hybrid Spatial Regions Philip Dougherty.
Principal components analysis (PCA) as a tool for identifying EEG frequency bands: I. Methodological considerations and preliminary findings Jürgen Kayser,
Clinical research data interoperbility Shared names meeting, Boston, Bosse Andersson (AstraZeneca R&D Lund) Kerstin Forsberg (AstraZeneca R&D.
Using decision trees to build an a framework for multivariate time- series classification 1 Present By Xiayi Kuang.
Identifying “Best Bet” Web Search Results by Mining Past User Behavior Author: Eugene Agichtein, Zijian Zheng (Microsoft Research) Source: KDD2006 Reporter:
Atlas Interoperablity I & II: progress to date, requirements gathering Session I: 8:30 – 10am Session II: 10:15 – 12pm.
Instance Discovery and Schema Matching With Applications to Biological Deep Web Data Integration Tantan Liu, Fan Wang, Gagan Agrawal {liut, wangfa,
AQUAINT Mid-Year PI Meeting – June 2002 Integrating Robust Semantics, Event Detection, Information Fusion, and Summarization for Multimedia Question Answering.
ICCV 2009 Tilke Judd, Krista Ehinger, Fr´edo Durand, Antonio Torralba.
Development of NeuroElectroMagnetic Ontologies (NEMO): A Framework for Mining Brainwave Ontologies Dejing Dou 1, Gwen Frishkoff 2, Jiawei Rong 1, Robert.
Department of Geography Jeon-Young Kang · Yi Yang
Data Mining Jim King.
Theoretical and empiral rationale for using unrestricted PCA solutions
Waikato Environment for Knowledge Analysis
Machine Learning for Visual Scene Classification with EEG Data
NON-NEGATIVE COMPONENT PARTS OF SOUND FOR CLASSIFICATION Yong-Choon Cho, Seungjin Choi, Sung-Yang Bang Wen-Yi Chu Department of Computer Science &
Process Wind Tunnel for Improving Business Processes
Presentation transcript:

Feb 28, 2010 NEMO data meta-analysis: Application of NEMO analysis workflow to consortium datasets (redux)

Overview of NEMO Project Aims Design and test procedures for automated & robust ERP pattern analysis and classification Capture rules, concepts in a formal ERP ontology Develop ontology-based tools for ERP data markup Apply ERP analysis tools to consortium datasets Perform meta-analyses of consortium data Build data storage & management system

The three pillars of NEMO ERP Ontologies ERP Data ERP Database & portal Focus of this All-Hands Meeting Focus of this All-Hands Meeting

TODA Y TUTORIAL #2: Decomposition with PCA TUTORIAL #3: Segmentation with Microstates TUTORIAL #1: Viewing ERP Data in EEGLAB

TOMORROW TUTORIAL #4: Extracting ontology-based attributes And exporting to text or RDF

1.Collect ERP data sets with compatible functional attributes 2.Decompose / segment the ERP data into discrete spatio- temporal patterns for analysis & labeling 3.Mark-up patterns within each dataset: labeling of spatial & temporal characteristics (functional labels assigned in step 1) 4.Cluster patterns within data sets 5.Link labeled clusters across data sets 6.Label linked clusters (i.e., establish mappings across patterns from different dataset) Overview Steps in Meta-analysis

1.Collect ERP data sets with compatible functional attributes 2.Decompose / segment the ERP data into discrete spatio- temporal patterns for analysis & labeling 3.Mark-up patterns within each dataset: labeling of spatial & temporal characteristics (functional labels assigned in step 1) 4.Cluster patterns within data sets 5.Link labeled clusters across data sets 6.Label linked clusters (i.e., establish mappings across patterns from different dataset) Focus of 1 st Annual All-Hands Meeting

1.Collect ERP data sets with compatible functional attributes 2.Decompose / segment the ERP data into discrete spatio- temporal patterns for analysis & labeling 3.Mark-up patterns within each dataset: labeling of spatial & temporal characteristics (functional labels assigned in step 1) 4.Cluster patterns within data sets 5.Link labeled clusters across data sets 6.Label linked clusters (i.e., establish mappings across patterns from different dataset) Overview Steps in Meta-analysis

Combining Top-down and Bottom- up Methods for ERP Pattern Classification Gwen Frishkoff University of Pittsburgh Robert Frank, Haishan Liu, & Dejing Dou University of Oregon

Human Brain Mapping Current Challenges – Tracking what we know  Ontologies – Integrating knowledge to achieve high-level understanding of brain–functional mappings  Meta-analyses Important Considerations – Stay true to data (bottom-up) – Achieve high-level understanding (top-down) “Understanding without data is empty. Data without understanding are blind”

ERP Patterns 1,000 ms TIME (in 10s of ms) SPACE (Scalp Topography)

Superposition of ERP Patterns

What do we know? Observed Pattern = “P100” iff  Event type is visual stimulus AND  Peak latency is between 70 and 160 ms AND  Scalp region of interest (ROI) is occipital AND  Polarity over ROI is positive (>0) FUNCTION TIME SPACE ?

Why does it matter? Robust pattern rules would provide a good foundation for–  Development of ERP ontologies  Labeling of ERP data based on pattern rules  Cross-experiment, cross-lab meta-analyses

TOP-DOWN

BOTTOM-UP

Top-down vs. Bottom-up Top-DownBottom-Up PROS Familiar Science-driven (integrative) Formalized Data-driven (robust) CONS Informal Paradigm- affirming? Unfamiliar Study-specific results?

Combining Top-Down & Bottom-Up

A Case Study 1.Simulated ERP datasets 2.PCA & ICA methods for spatial & temporal pattern analysis 3.Spatial & temporal metrics for labeling of discrete patterns 4.Revision of pattern rules based on mining of labeled data

Simulated ERPs (n=80) P100 N100 N3 MFN P300 + NOISE

BOTTOM-UP

Pattern Analysis with PCA & ICA

ERP pattern analysis Temporal PCA (tPCA) – Gives invariant temporal patterns (new bases) – Spatial variability as input to data mining Spatial ICA (sICA) – Gives invariant spatial patterns (new bases) – Temporal variability as input to data mining Spatial PCA (sPCA) ✔ ✔ Multiple measures used for evaluation (correlation + L1/L2 norms) X

BOTTOM-UP

Measure Generation A B T1 T2 S1 S2 Vector attributes = Input to Data mining (clustering & classification) CoP CoN ROI± Centroids Input to data mining: 32 attribute vectors, defined over 80 “individual” ERPs (observations)

BOTTOM-UP

Data mining Vectors of spatial & temporal attributes as input Clustering observations  patterns (E-M accuracy >97%) Attribute selection (“Information gain”) Figure 3. Info gain results for spatial ICA. CoP CoN ✔ ± Centroids Peak Latency

Revised Rule for the “P100” Pattern = P100v iff  Event type is visual stimulus AND  Peak latency is between 76 and 155 ms AND  Positive centroid is right occipital AND  Negative centroid is left frontal SPACE TIME FUNCTION

What we’ve learned Bottom-up methods result in validation & refinement of top-down pattern rules  Validation of expert selection of temporal concepts (peak latency)  Refinement of expert specification of spatial concepts (± centroids) Alternative pattern analysis methods (e.g., tPCA & sICA) provide complementary input to bottom- up (data mining) procedures

Simulated ERP Patterns “P100”“N100”“N3”“MFN”“P300”

Applying NEMO Functional Ontology Angela Laird BrainMap Jessica Turner BIRNlex (now part of Neurolex) CogPO

Adoption of BrainMap Classification Schema (modified for NEMO specifics*) Each experiment dataset classified based on: Subject contrasts — for first phase of NEMO, focus on healthy young monolingual adults (no contrasts) Condition contrasts — Which are defined with respect to stimulus, response, and instruction attributes * See NEMO_functional.low for details

Agreed to identify more auditory ERP data sets Agreed to restructure this hierarchy: primed as a special case of ”activated” (primed = within WM/activation span  SOA<10 sec AND prime-targ lag <3 ?

Datasets for NEMO Metaanalysis #1

Proposed pipeline for first NEMO meta- analysis

LP1 dataset (semantic contrast)– tPCA results (reconstructed ERP) projected into data space

LP1 dataset ( Lexicality Contrast )– tPCA results (reconstructed ERP) projected into data space

Following are some selected factors from LP1 tPCA results to illustrate some possible issues with factor ordering & retention after rotation…

LP1 dataset (tPCA Fac 1)– Lexicality Contrast

LP1 dataset (tPCA Fac 2)– Lexicality Contrast

Some Preliminary Conclusions Factor Retention may still be an issue for us collectively to explore – Unrestricted rotation vs. data reduction prior to rotation – For unrestricted path, what number to retain at end (after rotation)? – Also for unrestricted path, how to order factors at end (after rotation) We agreed to explore these issues, try to decide on final analysis pipeline by some date in near future (TBD…)