Information Technology Reference
In-Depth Information
quality indexes are incomparable), and also to compare the predicted output
indexes obtained within the limits of all models. At this stage of investigation it is
clear that fuzzy hybrid regression models have great values of determination
coefficient analogue in comparison with coefficient of determination of a classical
model.
Results of comparison of real output data with the output data obtained within
the limits of all models are summarized in Table 6.3.
To recognize fuzzy output values indexes (6.8) are used.
As one can see from Table 6.3, the model output data obtained based on the
fuzzy hybrid regression of model with fuzzy coefficients coincide with real output
data in 90 %. The model output data obtained based on the fuzzy hybrid regression
of model with definite coefficients, coincide with real output data in
80 %.
Table 6.3
Real and model output data
Model value of fuzzy regression
Item
No.
Initial
data
Model value of
classical regression
With definite coefficients
With fuzzy coefficients
1
2
2
3
2
2
3
3
3
3
3
2
2
2
2
4
4
3
4
4
5
5
4
4
4
6
3
3
3
3
7
4
5
4
4
8
3
4
3
3
9
3
3
3
3
10
4
4
4
4
Also, the model output data obtained based on the classical regression of model
coincide with real output data in 60 %.
Thus, the carried out analysis of quality classical linear and fuzzy linear hybrid
regression models (constructed within the limits of the information of educational
process) allows drawing a conclusion about advantages of the developed fuzzy
hybrid regression model, and reasonably recommend application of this model for
processing information of educational process and its reliable prediction.
