Research Article

Evaluation of Several Machine Learning Models for Field Canal Improvement Project Cost Prediction

Table 1

The statistical properties of the dataset selected for the training modeling phase.

Parameter CD
Parameter nameCost of FCIPDuration of FCIP constructionArea servedTotal length of PVC pipelineNumber of irrigation valuesConstruction yearGeographical zone
UnitLE/FCIPDayHectareMeterNumberYearZone

Mean353463.3076.3449.41813.838.182013.200.00
Standard error7539.450.781.2826.100.230.100.00
Median320292.5875.0045.90753.758.002014.000.00
Mode514778.0064.0051.00530.005.002014.000.00
Standard deviation113843.2411.7119.31394.033.521.480.00
Sample variance12960283390.67137.20372.83155258.8412.362.180.00
Kurtosis−0.221.90−0.07−0.242.72−0.13−1.18
Skewness0.771.040.750.600.94−1.05−0.58
Range518824.5069.0085.001956.4526.005.000.00
Minimum186825.9858.0019.00119.001.002010.000.00
Maximum705650.48127.00104.002075.4527.002015.000.00
Sum80589633.5117405.0011265.43185554.071866.02459010.000.00
Count228.00228.00228.00228.00228.00228.00228.00
Confidence level (95.0%)14856.261.532.5251.420.460.190.00