Transitioning Experiences with Army Geo Spatial Center (AGC)

1
Transitioning Experiences with Army Geo Spatial
Center (AGC)

• Pradeep Mohan
• 4th Year PhD Student

2
Overview Best Practices
Practice AGC CrimeStat Libraries 1.0 CrimeStat 3.2
Function Level Documentation v
Parameter sensitivity analysis v
Pattern Analysis with test datasets v
Well defined Test Cases v v
Unit Testing v
Integration Testing v v
Alpha/ Beta Testing v v
API Documentation v v
Testing on Multiple Platforms v
Multiple Data formats v v v
3
Cascading Spatio-temporal pattern discovery
T1
T2
T3
Assault(A)
C.4
A.2
B.2
A.3
Bar Closing(B)
C.1
B.1
Drunk Driving (C)
C.2
C.3
A.1
A.4
Cascading spatio-temporal pattern (CSTP)

• Partially ordered subsets of ST event types.

Bar Closing
Drunk Driving

• Located together in space.

Assault

• Occur in stages over time.

Stages Bar Closing, Assault , Drunk Driving,
Hurricane, Climate change etc.
Other Applications Climate change, epidemiology,
evacuation planning.
4
Project Cascade models for multi-scale pattern
discovery J.W Dr. J.A. Shine and Mr. J.P.
Rogers (AGC)
Cascade Pattern Discovery1 Source Code Matlab
2009b Test Case Crime Data, Parameters
Source code independent of Toolboxes
Toolbox Dependencies
.MAT Files (Test Cases)
Shape Files

1. Patterns
2. Performance bottlenecks
3. Bugs
4. Other issues like visual display

1. Performance analysis
2. Pattern analysis
3. Parameter sensitivity

Entire Process2 Months
AGC
1 Pradeep Mohan, Shashi Shekhar, James A.
Shine, James P. Rogers. Cascading spatio-temporal
pattern discovery A summary of results. In Proc.
of 10th SIAM International Data Mining (SDM)
2010, Columbus, OH, USA 2 J.A. Shine,
J.P.Rogers, S.Shekhar, P.Mohan. Cascade models
for multi scale pattern discovery An Extended
abstract. In USARMY ERDC Conference 2009,
Memphis, TN
5
Project Cascade models for multi-scale pattern
discovery J.W Dr. J.A. Shine and Mr. J.P.
Rogers (AGC)
Pattern Visualization
Pattern Data structure changes
Performance Enhancement
Faster Algorithms3
Fixing Bugs
Revised and Tested Code
Parallelizable
? MPI Support in Matlab
Toolbox independence
? Migration to C
AGC Requirements
Our Actions
3 (Ongoing) Pradeep Mohan, Shashi Shekhar,
James A. Shine, James P. Rogers. Cascading
spatio-temporal pattern discovery A summary of
results. (Journal Version)
6
CrimeStat

• A Spatial Statistics Program for the Analysis of
  Crime Incident Locations

7
Our Contributions

• Crime Stat Libraries 1.01
• Set of .NET components distributed by NIJ
• Crime Stat v 3.2
• Statistical Simulation functions for Spatial
  Analysis Routines
• Scalability to Large Datasets
• Self-Join Index2

1 http//www.spatial.cs.umn.edu/projects/crimest
at-pub/beta/ 2 Pradeep Mohan, Shashi Shekhar,
Ned Levine, Ronald E. Wilson, Betsy George, Mete
Celik, Should SDBMS support the join index ? A
Case Study from Crimestat. In Proc. of 16th ACM
SIGSPATIAL International Conference on Advances
in Geographic Information Systems (ACM GIS 2008),
California, USA,2008.
8
Project Crime stat Libraries 1.0 J.W Mr. Ron
Wilson (NIJ) and Dr. Ned Levine (Ned Levine and
Associates)

• Performance Tuning
• Outputs in several formats
• Beta Testing-I Feedback

• Documentation
• Visual Outputs
• Wrap up

• Spatial Analysis
• Spatial Interpolation
• Journey to Crime
• Distance Analysis
• Alpha Testing Feedback

Feedback

• Documentation
• Testing Framework
• Feedback updates

Feedback

• Core Components
• Spatial Description

• Documentation

Alpha Testing
Feedback
Vijay Gandhi1 Year
Beta Testing-I
Beta Testing-II
Chetan1 Year
NIJ Beta Testers
Pradeep Mohan1.5 Years
Entire Process2 .5Years

• Algorithm Descriptions
• Test Cases