Titles and Abstracts of MoCSSy Graduate Student Seminar Series

The following posters were presented at the 2nd annual MoCSSy symposium

The following posters (and presenters) won prizes in the poster competition

1. Using Network Analysis Tools to Assess System Level Variables in the Obesity System Map (Tommy Merth)
2. Social Influences in Community of Drug Users (Suraj Singh)
3. Examining HIV Transmission Rate through Sexual Contact with the Cellular Automata Model (Herbert H. Tsang)

A system dynamics model of body weight regulation and obesity

Ozge Karanfil - Ph.D. Candidate (Chronic Disease Systems Modelling Lab, Biomedical Physiology and Kinesiology, SFU)

The incidence of overweight is increasing across the world, and its adverse effects on health and economy have made obesity a major public health issue. Factors influencing the regulation of body weight have been under intensive investigation. The growing interest in body weight regulation has also culminated in the growth of simulation models that are employed as a tool to investigate this complex system, and as a means for evaluating hypotheses concerning the induction and maintenance of obesity.

The purpose of this modeling study is to develop a dynamic representation of our body weight regulatory system in normal and obese states, and to examine the interactions between the body composition and food intake regulation to see their effects on body weight maintenance. Considering the feedback complexity of the underlying structure and the different levels of factors involved (genetic, dietary, life-style, socio-economic), body weight regulation constitutes a suitable area for system dynamics simulation modeling. It is of considerable interest to identify the potential environmental and metabolic leverage points which determine our body weight trajectory and the adiposity at which our body weight is maintained. We aim to explore how these leverage points operate and to explain their potential role in explaining increasing rates of obesity.

A multi-relational approach to spatial classification

Richard Frank - Ph.D. Candidate (Computing Science, SFU)

Spatial classification is the task of learning models to predict class labels for spatial entities based on their features as well as the spatial relationships to other entities and their features. One way to perform classification on spatial data is to use a multi-relational database, by transforming the spatial data into multi-relational data, and then applying Inductive Logic Programming (ILP) onto it. However this presents novel challenges not present in multi-relational data mining problems:

• the spatial relationships are implied only by their location and not explictly stated,
• the lack of analysis involving aggregation and neighbourhood properties,
• the time required for the classification task is too large, an efficient solution is required.
• The method is compared against existing methods by experimental evaluation on several real world crime datasets.

The attractiveness of city centers for crime

Richard Frank - Ph.D. Candidate (Computing Science, SFU)

Crime Pattern Theory predicts that the journey to crime, the offender's direction of travel to the criminal event, will normally coincide with some frequently travelled path, such as to work or shopping. In this paper we experimentally analyze this theory by looking at the directional component of offenders' journeys to crime in communities across all of British Columbia. On the individual level, we first calculate the direction 25,376 offenders took from their home locations to the sites of their crimes. We then analyze these criminal activity vectors to identify patterns in the journey to crime. Findings indicate that people living in suburbs strongly tend to journey towards the central city to commit their offenses. Analysis of the patterns within cities indicated that some cities exhibit an easily identifiable attractor toward which nearly all criminal activity flows while other cities exhibit weaker patterns with criminal activity vectors pointing toward many small and diffuse attractors rather than some single central attractor.

A Computational Framework to Simulate Human and Social Behaviors

Andrew Park, Mengting Sun, and Herbert H. Tsang (MoCSSy, SFU)

The main objective of project GENIUS is to develop a decision support, response planning and risk assessment system. As part of this system, we have constructed a parameterized human's movement modelling framework. This framework allows the exploration of spatial-temporal features of the environment and perform risk analysis through studying different potential scenarios. This system consist of human movement simulation and visualization components. By applying computer animation and simulation techniques to model virtual crowds of autonomous agent's movement within an environment, this flexible framework can be applied to many different areas, such as site planning, entertainment, training, human factors analysis, and egress analysis.

The Prevalence of Menstrual Pain and Associated Rist Factors among Iranian Women

Mahkam Tavallaee - MPH Student (Health Sciences, SFU)

Objectives: To estimate the prevalence of dysmenorrhea in Iranian women and investigate associated risk factors.

Methods: In a cross-sectional study in Tehran, Iran, 381 women (81% response rate, age 16-56 years) completed a questionnaire regarding dysmenorrhea in 2007. Descriptive statistics, spearman correlation, and ordinal logistic regression models were used. Confounding and effect-modification were explored for each association.

Results: The prevalence of no, mild, moderate, and severe menstrual pain was 10%, 41%, 28%, and 22% respectively. Age, socioeconomic status, fruit and vegetable intake, fatty diet, stress level, and family history of dysmenorrhea were all associated with menstrual pain. BMI, parity, smoking, and physical activity were not associated with dysmenorrhea.

Conclusion: Menstrual pain is a common complaint in Iranian women. The inverse association between fruit and vegetable intake and dysmenorrhea needs to be further explored and considered in terms of recommendation to reduce dysmenorrhea.

System Dynamics Model of the Homeless Population Flow in Tri-Cities

Andrew Park, Gema Plaza-Martinez, …zge Karanfil, Yanchao Wang and Hengameh Vahabzadeh

Homelessness is one of the major urban issues as the social and economic circumstances become unfavourable to many people. Many social problems such as lack of housing, substance abuse, mental illness, criminal activities, and victimization are caused by or related to homelessness. The various services such as housing, health care, policing, and community are playing important roles to resolve these problems. In order to understand how these services work together as a whole system for the homeless and their issues, we have launched a joint research project with Fraser Health Authority, RCMP, Hope for Freedom Society, and other agencies. We have developed the model of the flow of the homeless population among different services in Tri-Cities using system dynamics modelling technique. By simulating, testing, and modifying the model, we try to grasp how the whole system works and find a way to improve the system.

Using Network Analysis Tools to Assess System Level Variables in the Obesity System Map

T. Merth, C. Matteson, …. Karanfil, H. Buhler, E. Chalubeyeu, D.T. Finegood (Chronic Disease Systems Modelling Lab, Biomedical Physiology and Kinesiology, SFU)

Introduction: Complex systems science has been used extensively to address complex problems in areas as diverse as climate change and physiology. Recent efforts to take a systems approach to obesity have resulted in the production of an Òobesity system mapÓ by the UKÕs Foresight Programme. This system map has 108 variables (sorted into 8 clusters) and 304 direct connections between variables.

Methods: In its original form, the enormous amount of information contained in the map conveys the notion of complexity very well, but makes it difficult to get clarity on the details. To make sense of the information on relationships between variables, we partitioned the variables into their 8 clusters using the network analysis software Pajek.

Results: With only 8 variables, 35 connections between variables and the number of underlying connections represented in the thickness of the connecting arrows, several interesting trends were observed. First, there were some significant disconnects between sectors (e.g. food production strongly influenced food consumption, but consumption had no effect on production). Several sectors had strong, equal influences on each other (e.g. food production and social psychology). Finally, some sectors had a much greater degree of internal interaction (e.g. physiology) than others. We are currently examining the underlying connections that contribute to the dominance of specific cluster connections.

Conclusions: This analysis provides new insight into the causal links included in the Foresight ProgrammeÕs Obesity System map. By emphasizing the feedback loops between clusters, the map can be used to further our understanding of system level challenges to addressing obesity.

Social Influences Among Drug users and Mean Field

Afsaneh Bakhtiari - M.Sc. Student (Complex Systems Modeling Group & Mathematics, SFU)

Approximation of Cellular Automata Street-involved youth have a propensity for street-level illicit drug use and trade. There are approximately one million street-involved youth in the United States and about 150,000 in Canada. The need for a proper intervention and follow-up strategy seems remarkably clear. Mathematical modeling is a potentially valuable tool for developing such a strategy. We present a mathematical model of the influences among current and potential illegal drug users and traffickers. It is a stochastic cellular automata model. We study the phase diagram of the cellular automata model using both simulation and mean field analysis. The phase plane of the mean field and possible bifurcations of the system are explored. The mean field approximation of the model corresponds to a compartmental model. Mean field analysis typically provides a good picture of a cellular automata model near a bifurcation. The model allows us to compare the potential effectiveness of different types of responses to the drug epidemic. We find that both indirect and direct strategies are effective on their own, but combined strategies proved to be most effective

Examining HIV Transmission Rate through Sexual Contact with the Cellular Automata Model

Azadeh Alimadad, Vahid Dabbaghian, Suraj Singh and Herbert H. Tsang (MoCSSy, SFU)

People with risky sexual behavior are at much higher risk for contracting HIV infection. Social and behavioral influences, as well as not knowing the HIV status, can promote unsafe sexual contacts. Therefore in high-risk populations, such as a population which consists of sex workers, HIV has a tendency to spread more rapidly. In addition, research show that accessibility to a HIV test is an intervention method that targets such communities. In this project we propose a cellular automaton model to study the dynamics of the HIV epidemic in a community at high risk of contracting HIV. In the model we introduce the presence of influences that promote or discourage unsafe sexual contacts.

Trust Relationships Can Delay Market Stability

Bojan Ramadanovic, Warren Hare, Laurens Bakker and Warren Hare (MoCSSy, SFU)

Isaac Newton, after he had lost a staggering 20,000 pounds in the South Sea bubble, famously complained that he could ``calculate the motions of heavenly bodies, but not the madness of people.'' Despite this genius admitting defeat, the potentially lucrative venture of predicting (stock) market behaviour has attracted many researchers, but so far to no avail. Tempting as this approach may be, we have chosen to take a different approach because we believe a view of market behaviour as a complex social system may yield valuable new insights. Taking the Efficient Market hypothesis, which posits that all traders are omniscient and perfectly rational in their trading decisions, we have developed a model to see how trust among rational traders could influence price development. We found that a real-world trust network significantly delayed stabilisation. Interestingly, a randomly generated trust network could not reproduce this delay.

Effect of Sodium Reduction on Cardiovascular Diseases

Azadeh Alimadad - Ph.D. Candidate (Statistics & Health Sciences, SFU)

Cardiovascular disease (CVD) is the leading cause of death in adult Canadian. In 2002, CVD accounted for 74,530 Canadian deaths. According to Statistics Canada (1990-1994) CVD mortality rates rise gradually from west to east with the lowest death rate in British Columbia. Hypertension is the leading risk factor for mortality worldwide which is a major risk factor for (CVD). Evidence shows that reduction in blood pressure level achieved by reduction in sodium intake can be translated in decreased risk of stroke, myocardial infarction and congestive heart failure. I will present my preliminary results of estimating the effect of dietary sodium on cardiovascular events in Canada.

Social Influences in Community of Drug Users

Vahid Dabbaghian, Valerie Spicer, Suraj Singh (MoCSSy, SFU)

The spread of hard drug use has become a matter of serious concern in modern societies. Policy makers are searching for relevant strategies to effectively control drug addiction as well as the negative medical and criminal repercussions associated to drug use. The spread of hard drug use (cocaine and heroin) is ofÊ particular concern because this type of addiction can lead to high-risk practices such as needle sharing, crack cocaine pipe sharing and drug related criminal activities. Social interactions in a high-risk drug community play an important role in spread of social behaviours especially with those pertaining to drug use and drug acquisition. This research examines the concurrent spread of hard drug consumption and criminal behaviour using a mathematical approach called cellular automata.

Neighborhood Migration and Crime

Vahid Dabbaghian, Piper Jackson, Valerie Spicer, Katie Wuschke (MoCSSy, SFU)

Residential migration patterns are the result of complex processes involving households and their neighbourhoods. Simulation models assist in understanding this relationship by contextualizing residential mobility within a theoretical framework. The objective of this project is to model residential mobility resulting from the interaction between a change in the social structure of the household and the positive or negative attractors in the neighbourhood. Specifically, this study links residential mobility to the dynamic interplay between the micro-environment existing within a household and the meso-environment that structures a neighbourhood. The cellular automata model developed in this study incorporates transition rules which govern households when they form the decision to move.