PREDICTING THE 2013 SAINT LOUIS CITY HOMICIDE RATE SPENCER SCHNEIDENBACH SHAILESH LANJEWAR XUN ZHOU BEN HOLTMAN.

Slides:



Advertisements
Similar presentations
Annual growth rates derived from short term statistics and annual business statistics Dr. Pieter A. Vlag, Dr. K. van Bemmel Department of Business Statistics,
Advertisements

Indianapolis, Indiana Offender Notification Meetings.
The Nature of Crime and Victimization
Building pride in Cumbria Do not use fonts other than Arial for your presentations Cumbria Intelligence Observatory: Cumbria Atlas Rebecca Raw: Research,
Predicting Homicides By Group B. Our Source
In Search of Insights and Better Analytics Prepared for: St. Louis Metropolitan Police Department 1 Nikolay Filipets / Ankit Patel / Melis Yilmaz / Divya.
University as Entrepreneur A POPULATION IN THIRDS Arizona and National Data.
Homicide Prediction Robert Dobynes Ken Lynch Misty Schutzius Danelle Tate.
Saint Louis Police Department BI Implementation Proposal Ben Shailesh Spencer Will.
Group A: Dan Diecker, Uzair Bhatti, Puji Bandi, Latoya Lewis.
Changing Demographics in Texas
Ryan Buenemann Thomas Mcgeehon Birute Simkeviciute Thomas Starr Tam Tran Decision Support Systems for Business Intelligence 2011 HOMICIDE DETECTION.
Homicide Predictions : Saint Louis City IS6833 February 28, 2010 Group B: Chris Gaynor Robert Jones Kevin Prinke Poernomo Wikandjojo.
St. Louis City Crime Analysis 2015 Homicide Prediction Presented by: Kranthi Kancharla Scott Manns Eric Rodis Kenneth Stecher Sisi Yang.
The Nature & Method of Economics Chapter One. Definition of Economics Social science concerned with the efficient use of limited resources to achieve.
Marketing Research Information gathered for a specific need or want. Can be primary or secondary Lecture Dr. Geurts.
Quantitative Research
The Research Process. Purposes of Research  Exploration gaining some familiarity with a topic, discovering some of its main dimensions, and possibly.
Who is ProvPlan? Mission to promote the economic and social well-being of the city, its people, and its neighborhoods. 501(c)3 non-profit created in 1992.
National Public Health Performance Standards Local Assessment Instrument Essential Service: 1 Monitor Health Status to Identify Community Health Problems.
Integrating and Evaluating Multiple PSN Strategies in Chicago Trace Meares Andrew Papachristos University of Chicago Jeffrey Fagan Columbia University.
The Gender Gap in Educational Attainment: Variation by Age, Race, Ethnicity, and Nativity in the United States Sarah R. Crissey, U.S. Census Bureau Nicole.
1 Overview of Major Statistical Tools UAPP 702 Research Methods for Urban & Public Policy Based on notes by Steven W. Peuquet, Ph.D.
Who Attends Private Schools? Enrollment rates by ethnicity in California Magali Barbieri, Shelley Lapkoff, Jeanne Gobalet Lapkoff & Gobalet Demographic.
Crime Analysis St. Louis, MO Presented by: Kranthi Kancharla Scott Manns Eric Rodis Kenneth Stecher Sisi Yang.
Predicting 2013 Trend of Homicide Rate in St. Louis City By Shih-Hua Chen Yao Zhang Teng ma Group E URL:
The new HBS Chisinau, 26 October Outline 1.How the HBS changed 2.Assessment of data quality 3.Data comparability 4.Conclusions.
The Nature of Crime and Victimization Is crime really a significant problem? Is crime increasing or decreasing? Is crime becoming more serious? Where and.
Effects of Income Imputation on Traditional Poverty Estimates The views expressed here are the authors and do not represent the official positions.
Predicting Homicides in St. Louis City for 2013 Chad Iseman - Geoff Hickman - Jon Perkins - Yanhui Long - Mustafa Khalili.
Bobby Renaud SOC December 4,  When looking at violent crime and motor vehicle theft, do we see a relationship? If so how significantly ?
1 Immigrant Economic and Social Integration in Canada: Research, Measurement, Data Development By Garnett Picot Director General Analysis Branch Statistics.
Epidemiology The Basics Only… Adapted with permission from a class presentation developed by Dr. Charles Lynch – University of Iowa, Iowa City.
Uzair Bhatti Dan Diecker Puji Bandi Latoya Lewis IS 6833 ANALYTICS ASSIGNMENT PREDICTING HOMICIDE RATE IN ST. LOUIS CITY FOR 2013.
Recent Trends in Worker Quality: A Midwest Perspective Daniel Aaronson and Daniel Sullivan Federal Reserve Bank of Chicago November 2002.
St. Louis Homicide Analysis Nikolay, Melis, Divya, Ankit.
Planning for 2010: A Reengineered Census of Population and Housing Preston Jay Waite Associate Director for Decennial Census U.S. Census Bureau Presentation.
Sampling Methods. Definition  Sample: A sample is a group of people who have been selected from a larger population to provide data to researcher. 
1 Measuring Quality Issues Associated with Internal Migration Estimates Joanne Clements, Amir Islam, Ruth Fulton & Jane Naylor Demographics Methods Centre.
Additional analysis of poverty in Scotland 2013/14 Communities Analytical Services July 2015.
The Nature of Crime and Victimization Chapter 2.  Primary sources for measuring crime are:  Official Data (Uniform Crime Reports)  Victim Surveys (National.
Chapter 13: Correlation An Introduction to Statistical Problem Solving in Geography As Reviewed by: Michelle Guzdek GEOG 3000 Prof. Sutton 2/27/2010.
An Introduction to Statistics and Research Design
Copyright  2003 by Dr. Gallimore, Wright State University Department of Biomedical, Industrial Engineering & Human Factors Engineering Human Factors Research.
Data and Construction of Economic Table Washington Child Support Group December 2007 Session I.
Millennium Development Goals Uruguay vs. Tobago Created by: Talon Sweeten & Mandy Nelson.
Correlation & Regression Chapter 15. Correlation It is a statistical technique that is used to measure and describe a relationship between two variables.
Shashin Amatya Yi Gao Lauren Reuther INFSYS-6833 Group B Homicide.
PERFORMANCE MODELS. Understand use of performance models Identify common modeling approaches Understand methods for evaluating reliability Describe requirements.
 Looked at Regression Too Small Sample Size  Sought Correlations Too Many  Looked at Linear Trend Lines  Neighborhood Statistics What We Tried.
Demography The Study of Populations. What is Demography?  Gathering and analysis of information about human populations  Eg. birth rates, death rates,
American Educational Research Association Annual Meeting AERA San Diego, CA - April 13-17, 2009 Denise Huang Examining the Relationship between LA's BEST.
The Psychology of Prediction and Uncertainty Jason Baer.
Targeting of Public Spending Menno Pradhan Senior Poverty Economist The World Bank office, Jakarta.
An Analysis of the Geographic Incidence of Social Welfare Factors as they Relate to School Performance of Early Elementary School Children Purpose This.
Organization of statistical investigation. Medical Statistics Commonly the word statistics means the arranging of data into charts, tables, and graphs.
Developing A Thesis Chapter 2.1 – In Search of Good Data
Mapping for the Next Millennium How CrimeRisk™ scores are formed.
Greene County JCPC Gang Identification and Prevention Department of Juvenile Justice and Delinquency Prevention-Center for the Prevention of School Violence.
IS 6833 ANALYTICS ASSIGNMENT Ying Chen, Sri Murali, John Powell, Scott Weber.
Beginners statistics Assoc Prof Terry Haines. 5 simple steps 1.Understand the type of measurement you are dealing with 2.Understand the type of question.
Implications and Limitations The Asthma H.E.L.P. program demonstrates that an asthma management program can be integrated into the casework process of.
Relationship of vegetation to socioeconomic status in Austin, Texas Kimberly Nichter, Department of Geography and the Environment This study observes the.
Income, Education, & Ethnicity Erin Dawson Ball State University Geography 265.
The State of Fathers in the State of Hawaii by Selva Lewin-Bizan, Ph.D. Center on the Family, University of Hawaii and Hawaii State Commission on Fatherhood.
Examining Achievement Gaps
IV. Why Services Cluster Downtown Ch. 13 – Urban Patterns
An Introduction to Correlational Research
Correlation and Causality
Presentation transcript:

PREDICTING THE 2013 SAINT LOUIS CITY HOMICIDE RATE SPENCER SCHNEIDENBACH SHAILESH LANJEWAR XUN ZHOU BEN HOLTMAN

BACKGROUND Annual homicide rates for 157 large US cities Analyzed for 30 years – 1976 to 2005 Factors Resource deprivation/concentrated poverty Higher income inequality Higher percentage of divorced adult male population Higher unemployment rates Study in 30 nations Significant association between poverty and homicide Sources:

DIVERSITY Characteristics of neighborhoods Very significant in predicting homicide Conclusion: immigrant concentration unrelated or inversely related to homicide language diversity consistently linked to lower homicide 15 years of data ( ) St. Louis Homicide rate related to neighborhood characteristics Patterns differ according to homicide subtypes – general altercation, felony and domestic Sources:

NATIONAL GANG TRENDS Source: FBI 2011 National Gang Threat Assessment – Emerging Trends

MISSOURI GANG TRENDS Source: FBI 2011 National Gang Threat Assessment – Emerging Trends

MISSOURI GANG TRENDS  Seems to be lower than other states with only 0-2 members per 1000 people  Rise in gang “promotional teams”  Increased gang use of social media directed towards youth  Presence small as it may be of 490 gangs according to the FBI Gang Threat Assessment Source: FBI 2011 National Gang Threat Assessment – Emerging Trends

DATA SELECTION PRINCIPLE Timeliness - Annually? Quarterly? Monthly? Sufficiency - Sample size – St Louis City, at least 5 years, the factors can have potential impact on criminal Level of detail or aggregation - Amount for reported criminal annually, criminal ratio distribute by district and possible influence factors such as poverty level, education attainment, population, Income etc Understandability - Readable for the crime data. Freedom from bias - How to avoid that? Keep it simple Decision relevance - How to determine the boundary? Geographical? How many factors are relevant to the criminal occurs geographically?

DATA SELECTION PRINCIPLE Comparability - Each city is individual case for analytic, avoid comparing the other cites’ data and cut off the data which influenced by abnormal factors. Reliability - We can not control, however there may be un-reported and un- detected crime which can influence the analysis Redundancy - Mulit-resources? Cost efficiency - Costs concern update data annually Quantifiability - Use Ratio level data Appropriateness of format - Which is the appropriate way to demonstrate

DIMENSIONALITY OF MODELS Representation -Reported crime Time Dimension -How much of the activity of decision environment is being considered Linearity of the Relationship -Determine if categorized data are linear or nonlinear Deterministic Versus Stochastic -Linear regression, Stochastic modeling Descriptive Versus Normative - Descriptive - used for prediction

DIMENSIONALITY OF MODELS Causality versus correlation -How to determine? - use criminal distribution graph and the other possible factor which has the positive or negative relate on them Methodology Dimension -Complete enumeration, algorithmic, heuristic simulations and analytical -Complete enumeration – large sample amounts required -Algorithmic – extremeness' value method -Heuristic - if math would not help -Simulation – external influence? Hard to identify -Analytical – speared parts for the whole process

MODELS WE CONSIDERED  Linear regression  Model based on Census, American Community Survey data  Predict crime based on population factors -Saint Louis Police Department Neighborhood Statistics -U.S. Census American Fact Finder Statistics -American Community Survey -Poverty Level -Educational Attainment -Lack of Core Family Stability Single Parent Families – Mothers with no husband present -Income -Race

MODELS WE CONSIDERED (CONT)  Linear Regression  Census data  Research only occurs once a decade  Hard to measure trends for predictions  American Community Survey  Broken down at the macro level (entire city)  Can’t measure by neighborhood, district  Conclusion:  Still useful for identifying problem areas inside a city  Best for a one-time “snapshot” to see what correlations exist and attack those problems  Largely outside the scope of what the SLMPD does

MODELS WE CONSIDERED (CONT)  Rolling average  Model based on past homicides  Weighs more recent data higher than other data  Pros  Data is easily accessible and accurate  Model is simple and pretty accurate  Cons  Does not predict big, one time events  Model data varies the more homicides are committed

MODELS WE CONSIDERED (CONT)  Rolling average  District vs. neighborhood  SLMPD uses districts  Most crime seems to be concentrated in several large areas  Districts it is  Quarters, months, years?  Years – too macro  Months – too micro – data is too wildly distributed  Measuring by quarters provides a nice balance between micro vs. macro and data accuracy

DECISION POINT  Rolling average it is  Regression model can’t be trended  Best model based on all available data

EXAMPLE MODEL  Rolling average it is  Best model based on all available data  Our model:  Prediction based on last 4 quarters  Last quarter: weighted by 0.4  2 nd last: 0.4  3 rd last: 0.1  4 th last: 0.1

MEASURING THE MODEL

OUR PREDICTIONS

CONCLUSION  Prediction is difficult

QUESTIONS?

SOURCES FBI 2011 National Gang Threat Assessment – Emerging Trends. services/publications/2011-national-gang-threat-assessmenthttp:// services/publications/2011-national-gang-threat-assessment Saint Louis Police Department Statistics - American Fact Finder - Saint Louis Homicide Map -