Blog
Forecasting With IBM SPSS
Introduction
Forecasting is widely sorted out for across various fields, and this is due to our curiosity to know what happens next. It can be used to identify threats and opportunities using historical data.
With forecasting we can get followings done:
• Estimating the number of employees, a call center will need each day.
• Predicting the amount of people who will be admitted into the emergency room.
• Infer the demand for gas or electricity providers.
• Predicting the number of passengers to commute to a particular location.
The time series feature for SPSS is called IBM SPSS Forecasting. A time series is a collection of data samples made by the regular monitoring of variables within a specific range of time.
In time series forecasting, a model is used to make predictions about the future based on previously observed data and the process calculates predictions for time series models using multivariate ARIMA (or transfer function models), univariate Autoregressive Integrated Moving Average (ARIMA), and exponential smoothing. For one or more dependent variable in spss, this technique contains an Expert Modeler that automatically determines and estimates the best fitting ARIMA or exponential smoothing model, instead of choosing an acceptable model by trial and error.
broadband1_data. This hypothetical data shows the number of customers to a national broadband service, broken down by region. The data file includes totals for 85 region four-year, monthly subscribers. The task here is to make predictions for the next three months for each of the 85 local markets in this business problems, we are going to utilize the Expert Modeler.
Data view or Scanning.
Knowing the characteristics of your data before creating a model is always a smart idea. Asking questions like are they seasonal fluctuations in the data? Even though the Expert Modeler will automatically select the best seasonal or non-seasonal model for each series.
Secondly, we frequently get quicker results by restricting the search to non-seasonal models when seasonality is absent from the data. We can obtain a general idea by showing the overall number of subscribers across all 85 local markets without looking at the data for each of them.
From the SPSS menus choose: Analyze >>> Forecasting >>> Sequence Charts as shown in figure 1.
Figure 1: Forecasting using Sequence Chart
Total Number of Subscribers should be Selected and drag it into the Variables list and then Select Date and drag it into the Time Axis Labels box. Click OK as shown in figure 2.
Figure 2: Selecting variable and timestamp for the sequenced chart.
Figure 3: A plot of Total numbers of Subscribers over time
Using sequenced chart to understand the market trend
In this blog, forecasting using IBM SPSS has been briefly introduced. Briefly in the sense that there is still more analysis that you can perform with this feature. Using the same data above, I challenge you to try out the function of ‘Autocorrelations’ and ‘Cross-Correlations’ under Forecasting feature.
Related blogs: Discriminant Analysis, Ratio Analysis
No comments added
Your one-stop website for academic resources, tutoring, writing, editing, study abroad application, cv writing & proofreading needs.