In data statistics, exponential smoothing is an analytical method that predicts future development trends based on the patterns of past time series, with the weight of data decreasing in chronological order.
How to operate SPSS index smoothing method?
In order to better explain the relevant knowledge of exponential smoothing method to everyone, below, I will demonstrate the specific operation steps of exponential smoothing method.
1. Import data. After entering the main operating interface of SPSS, click on the "File" tab in the upper left corner and use the "Open or Import Data" command to import the original file that needs to be analyzed.
Figure 1: Importing Data
2. Define the date and time. After the file is successfully imported, switch to the 'Data' tab and use the 'Define Date and Time' command within it.
Figure 2: Defining Date and Time
3. Define date settings. After entering the setting window for defining dates, we match the dates according to the text content. Here, I set the [case] type to year based on the year in the data.
Figure 3: Defining Date Settings
4. Sequence diagram. Then switch to the 'Analysis' tab and click on the' Time Series Prediction - Sequence Chart 'command in sequence.
Figure 4: Sequence diagram
5. Sequence diagram settings. After entering the settings interface of the sequence chart, drag the variable year to the "Variables" checkbox, drag the sales of the analysis object to the "Timeline Label" checkbox, and then click the "OK" command at the bottom.
Figure 5: Sequence diagram settings
6. View the results of linear analysis. In the pop-up result viewer, we can use the TS graph at the bottom to check if the text data follows a linear distribution. Once it does, we can perform exponential smoothing operations.
Figure 6: Viewing Linear Analysis Results
7. Create traditional models. Return to the main operating interface of SPSS, switch to the "Analysis" tab, and click on the "Time Series Prediction - Create Traditional Model" command.
Figure 7: Creating a Traditional Model
8. Time series modeler. After entering the time series modeler, we drag the sales amount to the dependent variable window and the year to the independent variable window. In the list of methods below, find and select the 'Exponential Smoothing' option.
Figure 8: Time series modeler
9. Condition setting. Click the 'Condition' button on the right and select the 'Non Seasonal Holt Linear Trend' option and the 'Dependent Variable Conversion - None' option in the model type. After completing the above settings, click the [Continue] and [OK] commands in sequence to start exponential smoothing analysis.
Figure 9: Condition Setting
Interpretation of SPSS index smoothing method results
After explaining the specific steps of the exponential smoothing method, I will now explain how to interpret the analysis results of exponential smoothing.
1. RMSE value. As shown in the following figure, slide to the "Model Design" column to view the "RMSE" value. This value represents the error between the actual value and the predicted value, and the smaller the value, the better the fitting effect.
Figure 10: RMSE value
2. Model fitting degree R squared. The validity of the data model can be determined by the R-squared value in the model fit chart. The closer the R-squared value is to 1, the higher the validity of the model, and vice versa.
Figure 11: Model Fit R-squared
3. Store results. After viewing the analysis results of index smoothing, we can click on the [File] tab and use the [Save or Save As] command to save the analysis results for future secondary viewing or derivative analysis.
Figure 12: Storage Results
