Nonlinear Modelling of Periodic Threshold Autoregressions Using Tsmars

Lewis, Peter A.W. and Ray, Bonnie K., "Nonlinear Modelling of Periodic Threshold Autoregressions Using Tsmars" . Journal of Time Series Analysis, Vol. 23, No. 4, pp. 459-471, 2002

Abstract:

We present new methods for modelling nonlinear threshold-type autoregressive behaviour in periodically correlated time series. The methods are illustrated using a series of average monthly flows of the Fraser River in British Columbia. Commonly used nonlinearity tests of the river flow data in each month indicate nonlinear behaviour in certain months. The periodic nonlinear correlation structure is modelled nonparametrically using TSMARS, a time series version of Friedman's extended multivariate adaptive regression splines (MARS) algorithm, which allows for categorical predictor variables. We discuss two methods of using the computational algorithm in TSMARS for modelling and fitting periodically correlated data. The first method applies the algorithm to data from each period separately. The second method models data from all periods simultaneously by incorporating an additional predictor variable to distinguish different behaviour in different periods, and allows for coalescing of data from periods with similar behaviour. The models obtained using TSMARS provide better short-term forecasts for the Fraser River data than a corresponding linear periodic AR model.

Keywods: Nonlinearity, periodic correlation, streamflows, threshold autoregression, TSMARS

Facts on IBM Software

Many businesses run multiple server farms but wish to consolidate them into a single smaller server farm. This is because most server farms are underutilizied or over provisioned. The boxes are typically running at 10% load which is quite costly and is not flexible. For example, one server farm goes hot and maxes out while the farm in the next room is still basically idle at 10%. XD allows administrator to define a single cluster (a node group) then monitor the workload and dynamically decide which boxes in the node group should host which application in order to meet these goals. If application A has a current response time of 1.5 seconds, XD will move resources away from applications B and C to increase the power dedicated to A and decrease its response time. XD can also predict that A will likely exceed its response time in 10 minutes based on a trend and react in anticipation of the event. This greatly simplifies the life of an administrator and allows the machines to be more efficiently used than a conventional multiple, independent farm of farms approach. XD also offers options to generate various email alerts when conditions are exceeded, it can restart servers when they appear to have a memory leak or after X requests.

Lotus's headquarters in Cambridge used to be divided into two buildings, the Lotus Development Building (LDB) (on the banks of the Charles River) and the Rogers Street building, located adjacent to the CambridgeSide Galleria. However, in 2001, then President and General Manager, Al Zollar decided not to renew the lease of LDB. The subsequent migration of employees across the street (and into home offices) generally coincided with what was probably the final exodus of employees from the company.

With Rational, IBM now will provide a complete software development environment for companies that want to integrate their business processes and software infrastructure across their companies and with suppliers, customers and employees. This requires the integration of data management, systems management, collaboration, transaction and business process integration middleware and software development. The acquisition of Rational will strengthen IBM’s leadership in each of these segments, including software development, and reinforces IBM's commitment to open industry standards. The acquisition is an important element of IBM's e-business on demand strategy.

This site is growing and will contain info on subjects like lotus notes, websphere application development and rational choice theory.