Anna T. Lawniczak

Prof. Dr. Anna T. Lawniczak

Department of Mathematics and Statistics, University of Guelph, Guelph, ON N1G 2W1, Canada
orcid

Research Focus & Metrics

Academic Career Length

23Years (2004-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 March 2006 in The European Physical Journal B
Verified
Journal Article
Closed access
PDF file
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
XTX. Tang
X. Tang

Verified
Journal Article
Open access
PDF file
Shengkun XieShengkun Xie
Shengkun Xie
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
Predictive modeling is a key technique in auto insurance rate-making and the decision-making involved in the review of rate filings. Unlike an approach based on hypothesis testing, the results from predictive modeling not only serve as statistical evidence for decision-making, they also discover relationships between a response variable and predictors. In this work, we study the use of predictive modeling in auto insurance rate filings. This is a typical area of actuarial practice involving decision-making using industry loss data. The aim of this study was to offer some general guidelines for using predictive modeling in regulating insurance rates. Our study demonstrates that predictive modeling techniques based on generalized linear models (GLMs) are suitable in auto insurance rate filings review. The GLM relativities of major risk factors can serve as the benchmark of the same risk factors considered in auto insurance pricing.

1 April 2007
Verified
Conference Paper
Closed access
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
PLPietro P. Lio`
Pietro P. Lio`
Shengkun XieShengkun Xie
Shengkun Xie
JXJiaying Xu
Jiaying Xu
Phase transition phenomena between non-congested and congested phases in packet traffic have been observed in many packet switching networks (PSNs). Using the PSN model we investigate the nature of fluctuations in number of packets in transit from their source to their destination, when the mean flow density into the PSN model is close to the phase transition point. A meaningful parameter of PSN behaviour near this critical point is the Hurst exponent that when larger than 0.5 is revealing of a long memory process, i.e. a fractional Brownian motion. In this paper we have used Hurst exponents and long range dependence to analyse PSN model behaviour. We have found that the DFA analysis and several methods for estimating the Hurst exponent suggest the presence of a long memory process for the PSN model using adaptive routing. However, we have not observed this in the case of static routing for the same type of incoming traffic. Thus, the packet traffic is more correlated in PSN model with adaptive routing than the static one. We present our finding, outline the work underway and discuss its expansion.

11 August 2023
Verified
Conference Paper
Open access
Shengkun XieShengkun Xie
Shengkun Xie
JZJin Zhang
Jin Zhang
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
Learning from imbalanced data has been an ongoing hot research area. By applying techniques for handling imbalanced data, machine learning or statistical models can significantly improve their prediction performance and mitigate bias, leading to more reliable and unbiased results. Data used to predict the fatality rate of car accidents is derived from various sources, including information at the person, vehicle, and collision levels. These data are typically imbalanced, and studying this type of data is highly desirable in improving road safety. Also, predicting a fatal event is crucial for better management and allocation of limited health resources. This study explores the impact of imbalanced data handling techniques on linear statistical models.The study illustrates the significant specificity improvement when imbalanced data is appropriately managed. The findings of this study provide valuable guidelines for health resource management, illuminating the influence of data imbalance on prediction accuracy and offering insights to improve the performance of predicting auto collision fatalities.

1 January 2006
Verified
Conference Paper
Closed access
BSBruno N Di Stefano
Bruno N Di Stefano
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
We study the movement of an autonomous roving object (ARO) having no knowledge of its territory, very limited memory, and limited intelligence. The ARO lives at discrete time steps in a two-dimensional universe consisting of discrete cells. The mission of the ARO is to cover its entire universe, without leaving any cell un-visited. We study the degree of success, that is, the percentage of territory that is guaranteed to be visited within a certain number of time steps, with "hundred percent" being full success. We consider different AROs, each characterized by different intelligence, each implemented using a specific finite state machine. We provide a very short review of and comparison with some existing AROs in the literature. Our simulation shows that introducing pseudo-randomness improves the performance of the ARO

18 September 2019 in Computation
Verified
Journal Article
Open access
PDF file
Shengkun XieShengkun Xie
Shengkun Xie
Anna T. LawniczakAnna T. Lawniczak
Anna T. Lawniczak
CGChong Gan
Chong Gan
For a better understanding of the nature of complex systems modeling, computer simulations and the analysis of the resulting data are major tools which can be applied. In this paper, we study a statistical modeling problem of data coming from a simulation model that investigates the correctness of autonomous agents’ decisions in learning to cross a cellular automaton-based highway. The goal is a better understanding of cognitive agents’ performance in learning to cross a cellular automaton-based highway with different traffic density. We investigate the effects of parameters’ values of the simulation model (e.g., knowledge base transfer, car creation probability, agents’ fear and desire to cross the highway) and their interactions on cognitive agents’ decisions (i.e., correct crossing decisions, incorrect crossing decisions, correct waiting decisions, and incorrect waiting decisions). We firstly utilize canonical correlation analysis (CCA) to see if all the considered parameters’ values and decision types are significantly statistically correlated, so that no considered dependent variables or independent variables (i.e., decision types and configuration parameters, respectively) can be omitted from the simulation model in potential future studies. After CCA, we then use the regression tree method to explore the effects of model configuration parameters’ values on the agents’ decisions. In particular, we focus on the discussion of the effects of the knowledge base transfer, which is a key factor in the investigation on how accumulated knowledge/information about the agents’ performance in one traffic environment affects the agents’ learning outcomes in another traffic environment. This factor affects the cognitive agents’ decision-making abilities in a major way in a new traffic environment where the cognitive agents start learning from existing accumulated knowledge/information about their performance in an environment with different traffic density. The obtained results provide us with a better understanding of how cognitive agents learn to cross the highway, i.e., how the knowledge base transfer as a factor affects the experimental outcomes. Furthermore, the proposed methodology can become useful in modeling and analyzing data coming from other computer simulation models and can provide an approach for better understanding a factor or treatment effect.