Integrated Feature Selection of ARIMA with Computational Intelligence Approaches for Food Crop Price Prediction

Table 4

The results of MARS₁ for rice price forecasting.

Variable selection results		BF	Coefficient
Variable	Relative importance (%)
Intercept			542.484
	100.0	BF1 = Max (0, yt−12 − 367)	0.985
		BF2 = Max (0, 552.09 − yt−12)	−1.258
	13.1	BF3 = Max (0, yt−11 − 356)	1.539
		BF4 = Max (0, yt−11 − 472.48)	−1.880
		BF5 = Max (0, 585.95 − yt−11)	0.524
	12.1	BF6 = Max (0, yt−10 − 363)	−2.410
		BF7 = Max (0, yt−10 − 472.48)	1.783
		BF8 = Max (0, yt−10 − 543.14)	0.454
		BF9 = Max (0, 623 − yt−10)	−0.234
		BF10 = Max (0, yt−10 − 623)	0.346