Forecasting Coursework

Forecasting Coursework

2010

Name: Mustafa Mohamoud

ID:08039718

Non-Seasonal Data

Section 1: Introduction

Section 1 of the case study report contains a statistical analysis of non-seasonal time series data, as well as appropriate forecasts for such data. The methods used will be a moving average models, regression analysis, smoothing methods and finally ARIMA models. For each method several models will be produced and the “best” model will be chosen based on investigation using various statistical tools and visual observations.

1.1 Data

The data being analyzed is the monthly unemployment rate (represented in a code YCNI given by the stastic gov website ), in london between 1992 Apr and 2008 Aug as recorded by the Office for National Statistics (ONS). The following is a definition of a LFS:

“The Labour Force Survey provides estimates of both the unemployment level and the unemployment rate. It is the rate that is the best indicator, because it measures the proportion of the economically active population who are unemployed and so takes account of changes in the size of the population over time, as well as changes in the level of unemployment. The Labour Force Survey (LFS) measures unemployment.”

(Source: http://www.statistics.gov.uk/downloads/theme_labour/unemployment.pdf)

Unemployment is a count of jobless people who want to work, are available to work, and are actively seeking employment. The Labour Force Survey was carried out every two years from 1973 to 1983. In 1984 the United Kingdom adopted the ILO( International Labour Organisation - an agency of the United Nations) definition of unemployment in the LFS.Unemployment is calculated using data from the Labour Force Survey (LFS), so it is subject to sampling differences.

Fig 1: Time series plot of monthly Unemployment rate from Apr1992 to Aug 2008.

The time series on Fig 1 shows monthly Unemployment rate from Apr1992 to Aug 2008 and the data displays no seasonality but there is a...