Table 5: Confusion matrix GLM model fit on Kariki_Farm data
with interaction detection using RST
From the experiment, we see that the GLM model, modeled using the
detected interaction terms from the Rough Set theory method, performed
better in accuracy compared to the model without the detected
interactions. Seeing these interactions helped determine the essential
features necessary for predicting the weather, reduced the
dimensionality of the data, and ultimately improved the accuracy of the
GLM model. The table below shows these results: