Time Series Analysis System for Person-Job Match Data.

Abstract

A requirement was identified for a general time series analysis package that could be applied to Person-Job Match (PJM) data. This paper presents a time series analysis system that forecasts and analyzes monthly time series data. The system consists of three forecast and analysis techniques-the Interactive Forecast and Analysis (Shiskin) procedure, the Box and Jenkins Univariate procedure, and the X-11 Census II Forecasting method. The Box and Jenkins technique is one of the most accurate statistical approaches presently available. The X-11 forecast method, which is easier to apply, has proven to be perhaps the most effective method for medium range forecasting with seasonal data. The Interactive Forecast Analysis (Shiskin) procedure provides a proven seasonal forecast routine plus a detailed statistical analysis of the time series prior to forecasting. In addition, a Time Series Data Preparation System was developed which creates and maintains the PJM data series. It is designed to provide a generalized and flexible method of extracting user specified data elements for creating a time series. (Author)

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 1981
Accession Number
ADA107885

Entities

People

  • David K. Roberts
  • Robert N. Leslie

Tags

DTIC Thesaurus Topics

  • Data Science
  • Delphi Method
  • Information Science
  • Statistical Analysis
  • Time Series Analysis

Readers

  • Aerial Delivery - Logistics and Supply Chain Management.
  • Computational Modeling and Simulation
  • Statistical inference.