Research Article

Methodology for Estimating the Cost of Construction Equipment Based on the Analysis of Important Characteristics Using Machine Learning Methods

Table 11

Forecasted values using trained models for loaders.

Params of heavy machineryReal priceEstimated valuesTrained models
Estimated priceEstimated range

Brand-New Holland, model-C227, weight-3.72 tons, payload-1200 kg, country of operation-Belgium, bucket volume-0.79 tons, interior protection design-cabin, steering mode-tank steering, max unloading height-2.4 m, hours worked-2400, year of manufacture-201734,900$59,860$58,521−67,120$Linear regression
62,640$55,240−64,199$Polynomial regression
32,008$29,067−34,949$Decision tree
32,899$28,077−33,653$Random forest
33,865$29,275−40,523$Neural network

Brand-John Deere, model-650, weight-9.32 tons, payload-4500 kg, country of operation-Belgium, bucket volume-0.3 tons, interior protection design-enclosed, steering mode-tank steering, hours worked-5778, year of manufacture-201357,500$49,860$47,309−64,714$Linear regression
43,135$39,680−55,390$Polynomial regression
54,322$49,270−64,125$Decision tree
58,090$54,637−59,230$Random forest
56,445$53,675−58,528$Neural network

Brand-Kubota, model-KX040, weight-4.17 tons, payload-420 kg, country of operation-Belgium, bucket volume-0.17, interior protection design-cabin, steering mode-rubber tracks, max unloading height-5.41 m, hours worked-1200, year of manufacture-201951,000$39,653$30,205−44,632$Linear regression
38,205$35,840−48,489$Polynomial regression
53,236$49,767−52,480$Decision tree
50,800$48,685−52,135$Random forest
49,930$47,913−51,987$Neural network