Nicholas I. Sapankevych

Time Series Prediction Using Support Vector Machines: A Survey

Time series prediction techniques have been used in many real-world applications such as financial market prediction, electric utility load forecasting , weather and environmental state prediction, and reliability forecasting. The underlying system models and time series data generating processes are generally complex for these applications and the models for these systems are usually not known a priori. Accurate and unbiased estimation of the time series data produced by these systems cannot always be achieved using well known linear techniques, and thus the estimation process requires more advanced time series prediction algorithms. This paper provides a survey of time series prediction applications using a novel machine learning approach: support vector machines (SVM). The underlying motivation for using SVMs is the ability of this methodology to accurately forecast time series data when the underlying system processes are typically nonlinear, non-stationary and not defined a-priori. SVMs have also been proven to outperform other non-linear techniques including neural-network based non-linear prediction techniques such as multi-layer perceptrons.The ultimate goal is to provide the reader with insight into the applications using SVM for time series prediction, to give a brief tutorial on SVMs for time series prediction, to outline some of the advantages and challenges in using SVMs for time series prediction, and to provide a source for the reader to locate books, technical journals, and other online SVM research resources.

Constrained Motion Particle Swarm Optimization and Support Vector Regression for Non-linear Time Series Regression and Prediction Applications

Support Vector Regression (SVR) has been applied to many non-linear time series prediction applications [1]. There are many challenges associated with the use of SVR for non-linear time series prediction, including the selection of free parameters associated with SVR training. To optimize SVR free parameters, many different approaches have been investigated, including Particle Swarm Optimization (PSO). This paper proposes a new approach, termed Constrained Motion Particle Swarm Optimization (CMPSO), which selects SVR free parameters and solves the SVR quadratic programming (QP) problem simultaneously. To benchmark the performance of CMPSO, Mackey-Glass non-linear time series data is used for validation. Results show CMPSO performance is consistent with other time series prediction methodologies, and in some cases superior.

Open border CONOPS simulation

This paper presents a simulation-based approach for analyzing open border security systems and architecture. This approach extends conventional, stand-alone sensor-based and communications based modeling, into a fully automated and integrated border security analysis suite which includes border security strategy trade-offs as well as resource utilization and cost effectiveness estimation. The analysis of the Measures of Effectiveness (MOEs) is accomplished via a discrete event simulation of Border Protection assets and forces against various threat scenarios. The focus of this simulation is to estimate and assess the effectiveness of various Border Protection solutions (personnel, equipment, CONOPs) in three dimensions-probability of apprehension of intruders, percent resource utilization, and cost effectiveness. The analysis suite also provides the engineering analytical capabilities needed to provide efficient and reliable engineering trade studies over various sized geographic regions and various terrain effects (flat earth, mountainous, and maritime).