Regional Mathematics and Statistics Conference

Submitted Abstracts

Name: Arnav Gupta, Luke Ni
Institution: UNC Wilmington
Title: Utilizing Entropy and Volatility for Machine Aided Analysis of Atrial Fibrillation
Abstract: Atrial Fibrillation (A-FIB) is a heartbeat arrhythmia, characterized by an irregularity in the R-peaks, the crest of each heartbeat on an electrocardiogram (ECG). Symptoms do not always appear during a doctor’s visit, which can later become life-threatening due to strokes. It could be important to offer a preliminary diagnosis to prevent such severe events. This project seeks to provide machine learning models that can help in the detection of A-FIB using RR-intervals, the distance between two R-peaks, analyzed by entropy and volatility. Twenty-three subjects from the MIT-BIH A-FIB database were analyzed in collections of twenty-five RR-intervals, using Leave One Person Out Cross Validation (LOPO-CV). Volatility, an important feature in LOPO-CV, measured the dispersion of the RR-intervals in seconds. Shannon, Approximate, and Sample Entropy were also used to classify A-FIB. Hyperparameter tuning on all tree-based methods and some non-tree-based methods was conducted to improve accuracy. The results indicated that Sample Entropy was a strong predictor of A-FIB. Tree-based methods such as Random Forests and Cat Boosting had higher accuracies than non-tree-based models. These results imply that heart rate monitoring devices like Apple Watches, could be used to aid the preliminary diagnosis of A-FIB.

Name: Taran Puvvala, Shravan Selvavel, Arihant Singh, Venkata Vadlamudi
Institution: UNC Wilmington
Title: Machine Learning in Identifying Atrial Fibrillation Through RR-Intervals
Abstract: An atrial fibrillation (AFib) patient not actively undergoing an arrhythmic episode is hard to diagnose accurately. This hurdle necessitates real-time detection methods that can be used to constantly monitor someone who may be at risk. This project focused on detecting AFib in near real-time using electrocardiogram (ECG) readings and a novel stepping window method. First, the distance between R-waves, or the RR-intervals, in the MIT-BIH AFib dataset’s 23 ECG recordings were calculated. Multiple features explored in prior research were chosen for the classification of the RR-Intervals. Features were calculated for each “stepping window”—a subset of 4 RR-Intervals—while factoring in the features of previous windows. The motivation behind this approach was to allow the model to make more accurate predictions given a very small subset of RR-intervals. We attained a maximum accuracy of approximately 94% using a CatBoost classifier and a selected set of features. Thus, we conclude that the stepping window method of feature calculation can be used to classify small sets of RR-intervals as AFib. In the future, we hope to explore a wider range of classifiers (including neural networks) and features—including different methods of calculating features in our stepping window system.

Name: Jeffrey Pinarchick
Institution: Elon University
Title: Predicting Key Factors of NFL Contract Extensions
Abstract: While modern NFL players are often paid more than most college students make in their lives, over 16% of NFL rookies end up filing for bankruptcy by the age of 36. Rookie NFL players receive an initial contract that typically covers their first four to five years of playing professionally in the NFL. To continue playing after this, fourth-year players are eligible to negotiate a contract extension or to enter into free agency. This research analyzes the data on player statistics and contract extensions of NFL quarterbacks and to build a model to predict whether a quarterback is offered a contract extension and the size of the contract extension. The data was obtained from Over The Cap and includes a variety of player statistics such as Passing Yards, Touchdowns, and Passer Rating. Our models indicate that the average passing touchdowns, rushing touchdowns, and rushing attempts during pre-contract seasons are some significant player statistics to predict whether a quarterback will receive a contract extension, and average passing touchdowns is the most significant player statistic to predict the size of the contract extension.

Name: Xiaohuan Xue
Institution: UNC Greensboro
Title: Constructing Covariance Functions for Axially Symmetric Processes on Spheres
Abstract: Covariance functions are used to characterize dependency in spatial statistics and the construction of covariance functions is critical when performing prediction or “kriging”. In this talk, we will discuss the the construction of covariance models for axially symmetric processes on the sphere. We will first review some of the recent development in this area, and we propose our preliminary results, then conclude the presentation with a discussion of future work.

Name: Xuan (Kelsy) Fei
Institution: Wake Forest University
Title: Predicting the Most Probable Path for Tipping in Arctic Sea Ice
Abstract: The Arctic sea ice is anticipated to melt away to a seasonally or perennially ice-free state by the end of the century. The decline of Arctic sea ice has the potential to form a feedback loop, speeding up climate change and global warming. The manner and expected time in which the Arctic sea ice may melt is a major source of concern for us. In this project, we used the Eisenman and Wettlaufer Arctic energy balance model with white noise, which is a piecewise-smooth and periodically forced stochastic differential equation. We investigated the most probable transition path between the ice-covered state and the ice-free state. To do this, we derived the Euler-Lagrange equations from the Freidlin-Wentzell functional corresponding to the original energy balance model. We computed the solution to these equations by a gradient descent algorithm and then compared with Monte Carlo simulations. We found that the most probable path follows a fast forcing regime. We analyzed the average escape time and the expected transition time of the most probable path.

Name: Victor Munoz Ocaranza
Institution: UNC Greensboro
Title: The effects of density-dependent dispersal on an ecological landscape model with logistic growth and grazing.
Abstract: We study positive solutions to a steady state reaction diffusion model arising in ecology. In particular, we consider a one-dimensional landscape model with logistic growth, grazing and density dependent dispersal. We analyze both positive density dependent dispersal as well as negative density dependent dispersal on the boundary. We generate bifurcation curves for positive solutions and analyze their evolution as the strength of the dispersal varies.

Name: Torre Caparatta
Institution: UNC Greensboro
Title: Zeros of Fractional Derivatives of Polynomials
Abstract: We study behavior of fractional derivatives $p^{(\alpha)} (x)$ of polynomials of degree $n$. First we find the exact location of their zeros, and then we investigate the intriguing movement of these zeros and establish their convergence properties as $\alpha \to n$.

Name: Alabi Banjoko
Institution: University of Ilorin
Title: A Data Mining Genome-Based Algorithm For Optimal Gene Selection And Prediction Of Colorectal Carcinoma
Abstract: This study presents a method for optimal selection of gene subsets to enhance the non-clinical diagnostic classification and prediction of colorectal cancer using gene expression level of 40 tumour and 22 normal colon tissues for 2,000 gene expression profiles obtained with an Affymetrix oligonucleotide array. A Hybrid multi-objective Support vector Machine (SVM) feature selection and classification algorithm was employed to determine the Biomarker gene subsets that are highly statistically and clinically relevant to the 62 (tumour or normal) responses of the gene expression levels. The genes selection was done in two stages with the first stage using the Bayesian t-test to prune the non-informative genes and the second stage employed the multi-objective optimization method that allows sequential addition of genes for optimal determination of the pre-selected gene subsets. The SVM with RBF kernel (SVM_RBF ) was fitted sequentially to select the set of near-optimal genes that are correlated with the response class. The optimally selected gene subset yielded an accuracy of 90.1% on the test data that were never used in the building process of the algorithm. Also, an estimated average of 86.94%, 91.92%, 85.87%, 92.64% and 91.56% was obtained for Sensitivity, Specificity, Positive predictive value, negative predictive value and Cross-validated Area under the curve (CVAUC) respectively. Furthermore, the results obtained from the principal component analysis and the complete linkage hierarchical clustering indicated near-perfect discrimination of the two clinical response groups of the colorectal cancer status of the patients.

Name: Blaine Farrell
Institution: UNC Wilmington
Title: Analyzing Implications of Collaboration in Academia Research with Scholarly Big Data and Social Network Analysis
Abstract: This project is centered on the use of Academic Social Networks and Social Network Analysis to discover the impact of collaboration on the rank of authors in research and if there are disparities between fields of study. An analyzable dataset was formed from a Microsoft Academic Graph to allow for examination into what drives a research author’s success among sixteen fields of study. An entity’s rank was determined by the probability of importance in relation to other entities. After finding the most important factor to the success of a paper was its author rank, the given entities (paper, author, conference, and journal) were analyzed to evaluate an author’s position in networks or to discover patterns explaining rank. These findings were later broken down into fields of study to observe differences between each subgroup. The use of a network contributed to more meaningful features for author entities and a deeper look into the impact of relationships that may influence a research author’s ranking. This project seeks to find what causes an author to have a higher probability of importance, the value of collaboration for this importance, and the future promotion of meaningful collaboration in academia.

Name: Emmanuel Adesina
Institution: Federal University of Technology, Akure. Nigeria
Title: Solutions Of First And Second Order Ordinary Differential Equations: Differential Transform Technique
Abstract: In this work, the differential transform technique is used to obtain solutions to first and second-order ordinary differential equations. Differential transform is an alternative iterative method for obtaining Taylor series solutions of differential equations. This method saves the large computational work for problems with large orders compared to the Taylor series; Differential Transform Method reduces the size of the computational domain and helps solve many problems easily. This method is illustrated by solving first and second-order ordinary differential equations and their results are compared with the exact solutions. Also, results are presented in tables and graphs.

Name: Shay-Ann Audain
Institution: Grambling State University
Title: Fitting a Mortality Curve
Abstract: Mortality retains considerable relevance when studying the factors that influence the intrinsic biology of ageing as well as when studying life insurance context, which considers a life-contingent series of payments that are contingent on the survival of a specific individual or group of individuals. The mortality rate is an indicator helping to determine the amount of life insurance premiums. In order to understand the mortality curve, this project investigates the trends to fit the historical data from the U.S Social Security Administration which is based on the mortality experience of a population during 2019 as used in the 2022 U.S Social Security Administration Trustees Report. We analyze the mortality trends across all ages based on American male and female mortality.

Name: Joia Zhang
Institution: University of Washington, Seattle
Title: A Robust Optional Quantitative Mixture RRT Model
Abstract: In this study, we introduce an optional quantitative RRT model that combines the elements of both the Pollock and Bek (1976) additive RRT model and the Greenberg et al. (1971) unrelated question quantitative RRT model. We examine the utility of the proposed mixture model using a unified measure of efficiency and privacy introduced by Gupta et al. (2018). We also account for the lack of trust in RRT models. The results show that the mixture model outperforms the two component models.

Name: Zunaira Azam
Institution: Lahore College for Women University
Title: Unit Inverse Weibull Distribution with Properties, Estimation, and Application
Abstract: Unit Inverse Weibull (UIW) distribution is introduced by using the exponential transformation. The statistical properties of Unit Inverse Weibull distribution are studied such as Survival function, Hazard function, Cumulative hazard function, Quantile and Median Function, Shannon Entropy, Density of Order Statistics. The maximum likelihood (ML) method is provided to estimate the parameters of the proposed distribution and also obtain the fisher information matrix (FIM). Application of Unit Inverse Weibull distribution to a real data set shows a better fit than other distributions based on unit interval.

Name: Vigneswaran Madappan Chinnasami
Institution: University of South Carolina Salkehatchie
Title: Rotation and Reflection of Magic Square Type Sliding Games
Abstract: A magic square type sliding game is a puzzle that has pieces of numbers in a grid with exactly one empty space, in which each row, each column, and each of the two diagonals sum up to the same number. To play this game, we slide one piece at a time into the empty space and try to rearrange all numbers to increasing order from the left grid to the right grid in each row, and from the top row to the bottom row. In general, not all sliding games can be rearranged (solved) this way. In our study, we create a sequence from each of the magic square and use inversion number to analyze the sequence. We then can use the change of inversion number to explain how the rotation and the reflection impact the solvability of this type sliding games.

Name: Samuel Byers
Institution: Florida State University
Title: Detecting spatial dependence with persistent homology
Abstract: There are many applications for determining if there is spatial dependency in geospatial data. The typical method uses permutation-based hypothesis testing and assumes the data is randomly distributed with respect to location. We propose a new topological data analysis (TDA) method that uses persistent homology to detect cluster-like patterns in data. We compare our method to a classical tool, Moran’s I, on a variety of real and simulated datasets to find classes of data types where the results diverge.

Name: Whitney Bennett
Institution: UNC Greensboro
Title: Low sensitivity to group members inhibits cooperation on evolving multiplayer networks
Abstract: We model a mobile population with arbitrary group size interacting over an underlying spatial structure using a Markov movement model. In this model, individuals strategically choose, based on their exploration strategy and the overall group composition, to stay at their “home” location or to move to a neighboring location. Interactions between individuals then take place in the form of a public goods game. Regardless of strategy, each individual receives a payout while only those who interact pay a cost. However, the less individuals that interact, the lower the overall group prosperity. This work builds on Erovenko et al. (2019) which investigated the effect of network topology on the evolution of cooperation. In this project, we consider how much the awareness towards group composition matters in regards to the evolution of the cooperative interactive strategy. We accomplish this by varying the sensitivity to group composition as part of an individual’s exploration strategy. We find a low sensitivity to group composition inhibits cooperation independently of the network topology.

Name: Purvi Contractor
Institution: University of Texas at Dallas
Title: Proving Inequalities in Machine Learning via Calculus
Abstract: The aim of this project is to show how esoteric aspects of a first calculus course play a role in modern machine learning. We will focus on the so-called concentration inequalities. Concentration inequalities bound the probability of a random variable deviating from its mean in the non-asymptotic regime. Famous examples include the Hoeffding’s and Bernstein’s inequalities. Typically skipped in most textbooks, these proofs will utilize little known facts from calculus. The research objectives are to fill in the proofs of the Hoeffding’s and Bernstein’s inequalities. In addition to discussing these proofs, the presentation will describe the connections between concentration inequalities and PAC agnostic binary supervised learning algorithms as well as highlight results from a literature review that illuminates how these inequalities are used in various industries. Future research will focus on the applications of the Hoeffding’s and Bernstein’s inequalities to other machine learning algorithms.

Name: Kiran Sarfraz
Institution: Lahore college for women university, Lahore
Title: DETERMINANTS OF CHILD MALNUTRITION USING ANTHROPOMETRIC INDICES: AQUANTILE REGRESSION ANALYSIS OF MULTIPLE INDICATOR CLUSTER SURVEY (MICS), PUNJAB 2017-18
Abstract: Among several serious public health problems, child malnutrition has been continuously a major challenge for decades in Pakistan. It is related with child mortality and morbidity, severely affecting the survival and early development of children. Hence, it is essential to take acquainted of its associates and determinants.
The goal of the current study to explore relationship among socioeconomic, demographic community health facilities on child malnutrition. Quantile regression and Ordinary Least Squares technique have been used to find out determinants responsible for child malnutrition in Pakistan by using anthropometric indicators. The age of child, birth order, area, sex, division, wealth index quintile, parental age and parental education have significant relationship with the anthropometric indicators of child malnutrition of Punjab, Pakistan. The findings of this study can provide assistance in making health guidelines to improve child nutritional outcomes, helpful to address the stunting, wasting and underweight in children under five-years.
Keywords: socioeconomic, demographic, community health facilities.

Name: Urooj khan
Institution: Lahore College for Women University Lahore Pakistan
Title: Spatial analysis of short birth interval among ever married reproductive age women of Pakistan
Abstract: World health organization recommended that an optimal birth interval for healthy pregnancy is at least two years (24 months). Short Birth Interval (<24 months) is one of the main hurdles to attain Sustainable Development Goal 3.1, 3.2. Short birth spacing has also negative effect on efforts to reduce fertility in the highly populated countries. Pakistan is facing both problems i.e. high fertility rate and high infant mortality. Objective of current research is to study the short birth interval (SBI) of ever married women in Pakistan. Data from Pakistan Demographic and Health Survey (2017-18) was used for analysis. Sample comprised of women of child bearing age (15–49) who had non first birth in the last five years preceding the survey. Spatial analysis is done using ArcGIS 10.7, SaTScan software and GWR 4 software. The spatial distribution of SBI was clustered all over Pakistan. The high prevalence of SBI was observed in Punjab, South Sindh, and Azad Jammu Kashmir. Father’s education, maternal age, and mother’s age at first birth were determined as the significant predictors of SBI. Tailored reproductive health care services in the hot spots areas are recommended instead of the existing uniform approach across the country.

Name: Uzma Yaqoob
Institution: Lahore College for Women University
Title: PERFORMANCE OF EXTENDED EXPONENTIALLY WEIGHTED MOVING AVERAGE (EEWMA) CONTROL CHART IN THE PRESENCE OF MEASUREMENT ERROR
Abstract: Control charts are the most effective way that is used to certify that manufacturing progressions are stable. Measurement error (ME) may affect the capability of the control chart to distinguish the mean shift. In real-world applications, measurement error is a common distortion component that affects the results of a process. The objective of this study was to see how effective is the EEWMA control chart for monitoring the process mean when ME exists in the data set. The average run lengths (ARLs) were generated through simulations to execute the EEWMA control chart’s efficiency in the presence of measurement error. It was concluded that the existence of ME negatively effects the detection ability of EEWMA control chart. It was further observed that by increasing the number of measurements per unit lessens the effect of measurement error on the chart’s performance. It was also established that EEWMA performed better than the EWMA control chart in the presence of measurement error.
Key words:- EEWMA, Measurement Error, ARLs

Name: Ayesha Waseem
Institution: Lahore college for women university, Lahore, Pakistan
Title: Contextual factors influencing incomplete immunization in Pakistan: A cross-sectional study based on PDHS 2012-13 & 2017-18
Abstract: Immunization is one of the important health indicator as it saves millions of lives of children and reduce risk of getting disease. Globally 6.3 million children died under the age of five years due to vaccine preventable disease and Pakistan ranked third among countries with most unvaccinated children. Objective of this study was to determine the factors affecting child immunization in Pakistan. For current study data was taken from the two waves of nationally representative Demographic and Health Surveys of Pakistan (PDHS 2012-13 & 2017-18). An ever married women who had children aged 12-23 months were included in this study. Immunization status of children were used as an outcome variable. In order to determine the effect of different factors on incomplete immunization, multilevel binary logistic model was fitted using STATA V.15. The barriers identified in the complete immunization of children are poor health facilities for pregnant women, gender biasedness against female children, lack of facilities in different areas, illiteracy and no media exposure of mother. Provision of better health facilities, launch of awareness campaigns for complete immunization of children particularly focusing female children in the low literacy areas can be effective interventions to reduce incomplete immunization and to attain the sustainable development goal 3.

Name: Abeera Shakeel
Institution: Lahore College for Women University, Lahore
Title: Impact of Demographic Factors on Total Fertility Rates using Poisson Regression Model: Comparative Study based on South & Southeast Asian Countries
Abstract: In this study, the impact of level of education, area of living and wealth index was studied on the fertility rates. Analysis was carried out for South & Southeast Asian countries using Poisson Regression Model. Data was obtained from the Demographic and Health Surveys for eleven South & Southeast Asian Countries. The TFR of Afghanistan was the highest among all countries. Total fertility rates for Pakistan and Timor-Lesta had entered in transitional level.
Pakistan, Cambodia, India, Nepal and Philippines showed high TFR for women with no education. Nepalese women had shown substantial decline in fertility as level of education increased. Afghanistan stood at top for both urban and rural fertility as compared to other countries in the region. Total fertility rate pattern showed steep decline in fertility rate for majority of countries with increase in wealth index. In Bangladesh, Pakistan and Afghanistan teenage fertility was found more as compared to other countries.

Name: Hasan Haq
Institution: City, University of London
Title: Modelling the evolution of structured populations under row-dependent movement
Abstract: In recent years, classic evolution models have been extended to incorporate more realistic features. A recent series of papers have developed a new evolutionary framework including structure, multiplayer interactions and movement. However, so far movement distributions have only involved independent movement e.g., without herding behaviour. In this presentation we develop a model to investigate the evolution of such a population under movement distributions where individuals are influenced by the movement of others. The interactions between individuals are modelled using the multiplayer public goods game. By incorporating these new movement distributions into existing evolutionary models, we demonstrate that certain levels of aggregation benefit specific types of individuals. Moreover, by extending the home fidelity parameter which is the measure of preference individuals have for staying at their home vertex to any positive value, we investigate a general approach of novel models such as the wheel to a wider class of structures.

Name: Alishba Hamid
Institution: LCWU
Title: Data analysis, discharge classification and prediction of hydrological parameters for the management of Tarbela Dam, Pakistan by using machine learning algorithms
Abstract: Throughout the history, it has been observed that floods are one of the major causes of disasters and can cause severe injuries to human beings and property damages. Dam offers flood control through providing storage capacity to impound flood water. Tarbela dam is an earth-filled dam situated in Khyber-Pakhtunkhwa province of Pakistan. Due to highest volume of reservoir, the low-lying downstream areas of the reservoir are being affected by reservoir storage and spillway discharge. The objective of this study is to find those variables effecting the spillway discharge. For the effective management, modeling techniques would not only be cost effective but also help the water managers in predicting the future scenarios of water discharge. The trends of hydrological parameters such as variables selection, discharge classification in terms of spillway gates opening, and prediction methods for the forecasting of hydrological parameters at the Tarbela Dam have been analyzed by machine learning algorithms using WEKA software. It has been concluded that water level and outflow are critical parameters of discharge management and SVM and J48 techniques gave highest value of correctly classified percentages. For forecasting, Multi-Layer Perceptron has been found more accurate.

Name: Neil Pritchard
Institution: UNC Greensboro
Title: Coarse Embeddability of the Space of Persistence Diagrams and Wasserstein Space
Abstract: When applying machine learning and statistical techniques one implicitly requires that they are working in a Hilbert space. When working in metric spaces not naturally contained in Hilbert space one may then seek a mapping into Hilbert space with controlled distortion. One such family of mappings are called coarse embeddings. It remains an interesting open question whether the space of persistence diagrams and Wasserstein space on R^2 coarsely embed into Hilbert space for p<2. In talk I will examine the connection between these problems and show they are equivalent for certain values of p.

Name: Maimoona Nazir
Institution: –
Title: Topp-Leone Odd Moment Exponential Half Logistic-Log-Logistic Distribution: Properties and Application
Abstract: In this article, a generalized three-parameter lifetime distribution known as Topp-Leone
Odd Moment Exponential Half Logistic-Log -Logistic (TL-OMEHL-LL) distribution is pro-
posed. Some mathematical properties of proposed model are studied such as quantile function, moments, bonferroni curve, lorenz curve and renyi entropy are presented. Model parameters are estimated by using the maximum likelihood (ML) method. Application on real life data is done to check the performance of the proposed model.

Name: Warda Mukhtar
Institution: Lahore College For Woman University, LHR
Title: Applying the OLS method and Quantile Regression to assessing the household saving behavior in urban and rural areas of Pakistan. Evidence from PSLM/HIES 2018-19
Abstract: This study aims to examine household saving behavior in urban and rural areas of Pakistan.
The study obtained micro data from the Household Integrated Economic Survey (HIES) 2018-19 and the Pakistan Social Living Standards Measurements Survey (PSLM) 2018-19 conducted by the Pakistan Bureau of Statistics (PBS). The effect of socioeconomic, and demographic on household savings is explored by using a quantile regression model. The considered characteristics are income, family size, age, marital status, dependency ratio, region, job status, occupation, and education. The study provides estimates of household savings using the least squares method. It is found that there exists a strong relationship between household saving behavior and socioeconomic and demographic characteristics (R2 = 53%).
The analysis is performed by the Ordinary Least Squares (OLS) method and compared with estimates obtained from the quantile regression (QR) model.
The Quantile regression results demonstrate that the income, age, marital status of currently married and no. of earners have a positive significant impact on household saving behavior. Moreover, age, income dependency ratio, no. of earners, and family size have a negative significant effect on saving behavior in Pakistan.
Whereas It was observed that people who are currently married have more savings as compared to others in the 90h quantile but less saving effect in 25th, 75th, 80th, and 10th. Marital status of currently married has a significant positive association with all quantiles. No. of earners and dependency ratio has a positive significant and negative significant effect on the 25th quantile respectively, with higher savings of self-employed in the 90th quantile as compared to those who are not self-employed while lower savings in 75th and 80th quantile. In rural areas of Pakistan saving rate increased at the 75th quantile while decreasing at the 10th quantile.

Name: Pujita Sapra
Institution: UNC Greensboro
Title: Mitigating Lack of Trust in Optional Binary RRT Models Using a Unified Measure of Privacy and Efficiency
Abstract: In this work, firstly the non-optional mixture of the Warner and Greenberg binary models proposed in Lovig et al. [2021] is presented. We show how this model accounts for untruthful responses due to a lack of respondent trust in the traditional binary RRT models. Next, we examine an optional version of the Lovig et al. [2021] model and show, theoretically and empirically, that it is more efficient than the corresponding non-optional model.

Name: Sheeza Rasheed
Institution: Lahore College For Women University, Lahore, Pakistan
Title: Socio-demographic determinants of BMI among women of reproductive age in Pakistan: A Pooled Analysis from PDHS (2012-13 and 2017-18).
Abstract: The study explored the socio-demographic determinants of BMI among Pakistani women of reproductive age from 2012-2018 using Pakistan Demographic and Health Survey. The response variable for the analysis was Body mass index. Ordinary least square and quantile regression models were used for statistical analysis. The percentage of obese women was the highest and the percentage of underweight women was the lowest. Age, education, frequency of watching TV, wealth index, husband’s education, and region showed a positive effect on women’s body mass index, while age of women at first birth, women’s working status, gender of household head residence and region showed negative effect on women’s body mass index. Obesity prevalence increased significantly among Pakistani women of reproductive age between 2012 and 2018. Obesity was found to be a more serious problem compared to under-nutrition in Pakistani women. Development of effective low-cost strategies to address the increasing burden of obesity should be a high priority.

Name: Wenhao Shou
Institution: UNC Greensboro
Title: Kernel Density Estimation Using Additive and Multiplicative RRT Models and Potential Role of Auxiliary Information
Abstract: In 2002 Ahmad introduced the kernel estimation of the density curve of a sensitive variable based on multiplicative RRT models and provided some theoretical results. In this article, we propose a kernel density estimator in the context of additive RRT models, which are more commonly used in the field of survey sampling. A simulation study is presented to validate the theoretical results from the previous work of Ahmad, and also compare the performances of the density estimators based on the additive and multiplicative RRT models. Simulations show that the proposed density estimator using additive scrambling performs better than the one using multiplicative scrambling, and it allows more error in the bandwidth selection in kernel density estimation. Some extensions to optionality and the use of auxiliary information will also be mentioned.

Name: Hasti Garjani
Institution: Delft University of Technology
Title: Improving Treatment of Metastatic Cancers Through Game Theory and Dynamical Systems Theory
Abstract: We investigate whether in a extended model by Pressley et al, a low constant dose can stabilize the tumor at a viable tumor burden and maximize the patient’s quality of life. Here we include competition coefficients to this model, motivated by ample in-vitro and in-vivo studies demonstrating the presence of competition and cooperation among cancer cells which impacts their growth. When stabilization of the tumor is possible, the constant lower dose is calculated based on Stackelberg evolutionary game setting presented by Salvioli. In the Stackelberg evolutionary game model, the physician as the leader selects the treatment dose to optimize the patient’s quality of life if the viable stable cancer population can be reached. For the extended model with competition between different cell types, the Stackelberg solution and Nash solution are investigated. While maximum tolerable does leads to cancer progression we can have Nash and Stackelberg solutions at the stabilization region. We show that Stackelberg strategies lead to the best results in terms of patient’s quality of life, followed by Nash strategies. Our research demonstrates that game theory is a useful tool for optimizing treatment within the context of evolutionary therapies.

[1] M. Pressley, M. Salvioli, D. B. Lewis, C. L. Richards, J. S. Brown, and K. Staˇnkov´a, “Evolutionary Dynamics of Treatment-Induced Resistance in Cancer Informs Understanding of Rapid Evolution in Natural Systems,” Frontiers in Ecology and Evolution, vol. 9, p. 681121, Aug. 2021.
[2] M. Salvioli, Game theory for improving medical decisions and managing biological systems. PhD thesis, Politecnico di Milano, Milano, Italy, March 2020.

Name: William McCance
Institution: University of California: Santa Barbara
Title: Binary Randomized Response Technique (RRT) with Measurement Error
Abstract: In real-world surveys, measurement error is inevitable as the difference between the actual value of the variable being measured and its recorded value. Many authors in the field of Randomized Response Technique (RRT) have studied the impact of measurement error on quantitative RRT models, but there are no studies founded on binary RRT models. In this article, we propose a binary RRT model under measurement error based on the previous work of Warner (1965) and also introduce the modulo operation to define the measurement error. A simulation study is presented to validate the theoretical findings. Simulations show that the measurement error factor cannot be ignored when using binary RRT models, and the proposed estimator from the binary RRT model under measurement error performs well.

Name: Gleb Gribovskii
Institution: UNC Greensboro
Title: The multiplayer Hawk–Dove game on evolving random networks
Abstract: We model a mobile population of Hawks and Doves interacting in groups of arbitrary size over a network using a Markov movement model. Individuals choose strategically whether to remain at their current location or move to one of the neighboring locations, depending upon their exploration strategy and the composition of their group. We consider a range of network topologies such as random Erdos–Renyi graphs, Barabasi–Albert (preferential attachment) graphs, random regular graphs, and Watts–Strogatz (small world) graphs. We investigate how the network topology and the movement cost affect the evolution of the population of Hawks and Doves. This is a preliminary report on work in progress.

Name: Aqsa Rafique
Institution: Lahore College For Women University
Title: Comparison of different estimation methods
for Exponentiated Generalized Frechet Geometric distribution
Abstract: The comparative study of estimation method for Exponentiated Generalized Frechet Geometric distribution is presented via maximum likelihood estimators, least square estimators, weighted least square estimators, Cramer-von Mises estimators, Anderson-Darling estimators, right-tail Anderson-Darling estimators, left-tail Anderson-Darling estimators, and maximum product of spacing estimators. The behaviour of various estimation methods using different samples sizes by using mean square error and average bias. The potentiality of the distribution is presented by utilizing real data sets.

Name: Josh Slater
Institution: UNCG
Title: Persistent Homology through Image Segmentation
Abstract: The efficacy of topological data analysis (TDA) has been demonstrated in many different machine learning pipelines, particularly those in which structural characteristics of data are highly relevant. However, TDA’s usability in large scale machine learning applications is hindered by the significant computational cost of generating persistence diagrams. In this work, a method that allows this computationally expensive process to be approximated by deep neural networks is proposed. Moreover, the method’s practicality in estimating 0-dimensional persistence diagrams across a diverse range of image datasets is explored.

Name: Mehar ul Nisa
Institution: Lahore College For Women University
Title: A NEW GENERLIZED FORM OF ODD GUMBEL TYPE 2 DISTRIBUTION
Abstract: The purpose of this study is to introduce new continuous distribution by using DUS transformation. We derive new distribution from the G-family distribution known as Power DUS Gumbel distribution. Statistical measures of the Power DUS Gumbel Distribution (PDUSGD) including probability distribution function, cumulative distribution function the mean, moments, reliability function, quantile function, hazard function, survival function Renyi entropy and order statistics have been studied. The maximum likelihood method used to estimate the parameters of the Power DUS Gumbel distribution. The Power DUS Gumbel distribution was applied to two real-world data sets in order to contrast the outcomes of other widely used distributions.
Key Words: PDUSG, PDF, CDF, MGF, MLE

Name: Nandakishor Krishnan
Institution: ELTE, Budapest
Title: Genesis of Ecto-symbiotic features based on Commensalistic Syntrophy
Abstract: Eukaryogenesis and organellogenesis have been recognized as major evolutionary transitions and are subject to in-depth studies. In the endo-symbiotic theory, a eukaryotic cell is assumed to have evolved from an endo-symbiotic association between unicellular hosts and symbionts of different species (non-nucleated bacteria) which were once capable of independent existence. Acknowledging the fact that the initial interactions and conditions of cooperative behavior between free-living single-celled organisms are widely debated, we narrow our scope to a single mutation that could possibly have set off the transition to multi-species intimate associations. We investigate the ecological and evolutionary stability of inter-species associations with vertical transmissions based on uni-directional syntrophy. The dynamics of the population densities of the involved species are represented using a set of ordinary differential equations and the growth rates of each species are represented using novel Malthusian functions. We utilize the concept of evolutionary substitution, and the ecological fixation of the ecto-commensalistic association is mathematically modeled in terms of the local asymptotic stability (using linearization) of certain fixed points corresponding to the coevolutionary dynamical system.