Gary Weckman - Academia.edu (original) (raw)

Papers by Gary Weckman

— Large wireless network infrastructures experience concurrent or overlapping service outages due... more — Large wireless network infrastructures experience concurrent or overlapping service outages due to equipment and link failures. The frequency, duration, and impact of such episodes are of interest to users and network operators alike. Here, a research project which investigates through simulation the characteristics of concurrent network outages in large wireless network infrastructures is presented. The dependability attributes used to gain a perspective on this issue are network reliability, availability, maintainability and survivability. To assess these attributes in this setting, a new term, called an “impact epoch”, is introduced. Epochs are defined as single, concurrent, or overlapping outages in time, consisting of n different outages. A wireless network is expanded in size and epochs observed as the network grows. The new proposed metrics offer valuable insights into the management of restoration resources. Simulations proved invaluable in identifying multi-outage epochs,...

— This paper presents a comprehensive analysis of the reliability and survivability of US local t... more — This paper presents a comprehensive analysis of the reliability and survivability of US local telecommunication switches over a 14-year study period (from 1996 to 2009). Using local switch outage empirical data, the causes, failure trends and impacts have been identified, analyzed and assessed. A total of 12,860 switch outages were investigated for which very significant reliability growth was identified over the study period. Outages were also studied temporally, from time of day, day of week, and month of year perspectives. Additionally, 2,623 of the outages were found to come from only 156 unique switches, each of which experienced eight or more outages over the study period. The data were separated into two categories, for comparison: more frequently failing switches and less frequently failing switches. Major findings are that scheduled maintenance activities and hardware failures are the major causes of outages in local telecommunication switches; there are significant causa...

Prediction is the method of determining future values based on the patterns deduced from a data s... more Prediction is the method of determining future values based on the patterns deduced from a data set. This research compares various data mining techniques-namely, multiple regression analysis in statistics, Artificial Neural Networks (ANN), and the Group Method of Data Handling (GMDH), including both with and without feature selection. Currently, the literature suggests that GMDH, an inductive learning algorithm, is an excellent tool for prediction. GMDH builds gradually more complex models that are evaluated via a set of multi-input, single-output data pairs. ANNs are inspired by the complex learning that happens in the closely interconnected sets of neurons in the human brain and are also considered an excellent tool for prediction. This article is the beginning of a more detailed research project to investigate how well GMDH performs in comparison to other data mining tools.

As a network infrastructure expands in size, the number of concurrent outages can be expected to ... more As a network infrastructure expands in size, the number of concurrent outages can be expected to grow in frequency. The purpose of this research is to investigate through simulation the characteristics of concurrent network outages and how they impact network operators’ perspective of network dependability. The dependability investigated includes network reliability, availability, maintainability and survivability. To assess this phenomenon, a new event definition, called an “impact epoch”, is introduced. Epochs are defined to be either single, concurrent, or overlapping outages in time, which can be best assessed with new metrics and simulation. These metrics, Mean-Time-To-Epoch, Mean-Timeto Restore-Epoch along with percentage time the network is not in an epoch state (Quiescent Availability) and Peak Customers Impacted, are investigated. A case study based upon a variable size wireless network is studied to see what insights can be garnered through simulation. The new proposed met...

This paper presents a reliability analysis of local telecommunication switches experiencing frequ... more This paper presents a reliability analysis of local telecommunication switches experiencing frequent outages in the United States, based upon empirical data. Almost 13,000 switch outages are examined and over 2,500 are found to originate with just 156 switches experiencing eight or more outages each over a 14-year period. Telecommunication switch outage statistics are analyzed for this multiyear period, allowing examination into switch failure frequency, causes, trends, and impacts. Failure categories are created by reported outage cause codes, including human error, design error, hardware failure, and external factor causality categories. Principal findings are that there are significant differences in the switch and outage characteristics for switches experiencing more frequent outages/failures. Additionally, time series analysis indicates significant reliability/survivability deterioration in switches experiencing more frequent outages. Keywordstelecommunication; reliability; loc...

Journal of Industrial and Systems Engineering, 2010

Neural networks were used to estimate the cost of jet engine components, specifically shafts and ... more Neural networks were used to estimate the cost of jet engine components, specifically shafts and cases. The neural network process was compared with results produced by the current conventional cost estimation software and linear regression methods. Due to the complex nature of the parts and the limited amount of information available, data expansion techniques such as doubling-data and data-creation were implemented. Sensitivity analysis was used to gain an understanding of the underlying functions used by the neural network when generating the cost estimate. Even with limited data, the neural network is able produced a superior cost estimate in a fraction of the time required by the current cost estimation process. When compared to linear regression, the neural networks produces a 30% higher R value for shafts and 90% higher R value for cases. Compared to the current cost estimation method, the neural network produces a cost estimate with a 4.7% higher R value for shafts and a 5% ...

A 5.3% increase in motor vehicle traffic crashes in 2012 [1] brings up the discussion of related ... more A 5.3% increase in motor vehicle traffic crashes in 2012 [1] brings up the discussion of related traffic safety parameters. This paper considers the analytical assessment and evaluation of various highway safety factors that will eventually trigger fatalities. The related safety parameters are mainly divided into four categories-- economical investment, system usage, road condition, and personal safety. Three data mining algorithms-- K-nearest Neighbors algorithm (KNN), Random Forest and Support Vector Machine (SVM), and also a probabilistic Artificial Neural Network (ANN)-- are used for the prediction of highway fatalities among the eight different safety indicators. According to the Bureau of Transportation Statistics’ most recent available data, the analysis of this study covers the years from 2003 to 2011. The preliminary results indicated that out of the three, the proposed Random Forest data mining approach predicted the data with the highest percentage. The sensitivity analys...

Fierce competition in the market has forced companies to study their supply chain networks more. ... more Fierce competition in the market has forced companies to study their supply chain networks more. Due to increased social awareness and stricter governmental laws and legislation, the green closed-loop supply chain (GCLSC) has been reviewed more recently. The main goal of this study is to propose a multi-objective model for optimization of a comprehensive green closed-loop supply chain network with a multi-period multi-stage network, including the manufacturer, distributor, customer market, collection, recovery, and disposal centers under uncertain demand. To handle the uncertain parameter, we utilized chance constraint fuzzy programming. We considered different objective functions, consisting of maximizing income, minimizing total supply chain cost, and minimizing total CO2 emissions (i.e., CO2 emitted from facility centers and various transportation modes). Aimed at achieving optimal values, we utilized a carbon-pricing approach to transform the problem into a single objective func...

In this paper, we investigated three questions. First, to what extend do wine critic scores and d... more In this paper, we investigated three questions. First, to what extend do wine critic scores and descriptions influence consumer-buying decisions? Second, to what extent this influence varies with price? Third, how do demographics affect consumer decisions? The experimental design consisted of convenience samples from four different stores in a major United States (US) metropolitan market, with random assignment of consumers to different groups, who completed a total of 240 survey questionnaires. The dependent variable was likelihood to buy wine when presented with varying amount of wine critic information (a control group and three experimental groups with different levels of information). Independent variables included wine price, age, gender, wine interest, store type and location. Major findings include some surprises. For a 20bottleofwine,thecriticinformationwasnotafactoronlikelihoodtopurchase,whilecriticinformationwasafactorfora20 bottle of wine, the critic information was not a factor on likelihood to purchase, while critic information was a factor for a 20bottleofwine,thecriticinformationwasnotafactoronlikelihoodtopurchase,whilecriticinformationwasafactorfora50 bottle. These findings...

2001 Annual Conference Proceedings

International Journal of Logistics Systems and Management

Energies

As the level of greenhouse gas emissions increases, so does the importance of the energy performa... more As the level of greenhouse gas emissions increases, so does the importance of the energy performance of buildings (EPB). One of the main factors to measure EPB is a structure’s heating load (HL) and cooling load (CL). HLs and CLs depend on several variables, such as relative compactness, surface area, wall area, roof area, overall height, orientation, glazing area, and glazing area distribution. This research uses deep neural networks (DNNs) to forecast HLs and CLs for a variety of structures. The DNNs explored in this research include multi-layer perceptron (MLP) networks, and each of the models in this research was developed through extensive testing with a myriad number of layers, process elements, and other data preprocessing techniques. As a result, a DNN is shown to be an improvement for modeling HLs and CLs compared to traditional artificial neural network (ANN) models. In order to extract knowledge from a trained model, a post-processing technique, called sensitivity analysi...

Advances in Artificial Neural Systems

Artificial neural networks (ANNs) are powerful empirical approaches used to model databases with ... more Artificial neural networks (ANNs) are powerful empirical approaches used to model databases with a high degree of accuracy. Despite their recognition as universal approximators, many practitioners are skeptical about adopting their routine usage due to lack of model transparency. To improve the clarity of model prediction and correct the apparent lack of comprehension, researchers have utilized a variety of methodologies to extract the underlying variable relationships within ANNs, such as sensitivity analysis (SA). The theoretical basis of local SA (that predictors are independent and inputs other than variable of interest remain “fixed” at predefined values) is challenged in global SA, where, in addition to altering the attribute of interest, the remaining predictors are varied concurrently across their respective ranges. Here, a regression-based global methodology, state-based sensitivity analysis (SBSA), is proposed for measuring the importance of predictor variables upon a mode...

Environmental monitoring and assessment, 2017

Demand forecasting plays a vital role in resource management for governments and private companie... more Demand forecasting plays a vital role in resource management for governments and private companies. Considering the scarcity of water and its inherent constraints, demand management and forecasting in this domain are critically important. Several soft computing techniques have been developed over the last few decades for water demand forecasting. This study focuses on soft computing methods of water consumption forecasting published between 2005 and 2015. These methods include artificial neural networks (ANNs), fuzzy and neuro-fuzzy models, support vector machines, metaheuristics, and system dynamics. Furthermore, it was discussed that while in short-term forecasting, ANNs have been superior in many cases, but it is still very difficult to pick a single method as the overall best. According to the literature, various methods and their hybrids are applied to water demand forecasting. However, it seems soft computing has a lot more to contribute to water demand forecasting. These cont...