Estimation of Incremental Haulage Costs by Mining Historical Data and Their Influence in The Final Pit Definition PDF
Estimation of Incremental Haulage Costs by Mining Historical Data and Their Influence in The Final Pit Definition PDF
Estimation of Incremental Haulage Costs by Mining Historical Data and Their Influence in The Final Pit Definition PDF
• bench. where
• destination. a is the actual attribute value and
• lift (Delta Z). b(m) is the predicted value using the model (m).
Chen (2001) suggests dividing the data into three Equation (1) assumes a normal distribution to com-
subsets: training, testing and evaluation sets for predic- pute the probabilities for both predicted and actual
tive modeling. For this analysis, two data sets were ar- values. Table 2 summarizes the different scores achieved
ranged: a training set to build the model (e.g., to define for the predictable attributes (potential cost drivers), for
equations between input/predictable attributes) and both training and testing data sets. Additionally, Table 2
a testing set to evaluate its accuracy. A data mining includes the square of the correlation coefficient (R2)
technique called Decision Trees was chosen for model for the models. It can be noted clearly that “haultime”
construction. For the case of continuous variables, the achieved the highest score values compared to the other
decision tree algorithm uses linear regression to decide indicators in both data sets. Figure 4 shows the decision
where a decision tree splits. A regression formula is cre- tree for “haultime,” with a regression formula in each
ated in each node of the tree. A split occurs at a point node of the tree. For instance, a haul profile having the
of nonlinearity in the regression formula. Generally, crusher (CR) as destination with a bench elevation less
a regression formula contains one or more regressors than 2,564 has the following expression:
(input attributes). If no regressor is present in the for-
mula, the result tree contains a constant in each leaf Haul time = 8.808 + 0.024* (2)
node. (Delta Z - 122.071) – 0.024*(Bench – 2,541.929)
Eight destinations were considered for model
construction according to the material type as-
signed. Twenty benches, with elevations between
Table 2
2,510 and 2,770 m (8,230 to 9,090 ft) above sea Scores and R2 for training and testing data sets.
level, were as well considered as origin points. Training set Testing set
Figure 2 shows a dependency network, which is Attribute name Units Score R2 Score R2
used generally as an exploratory data analysis.
This tool is included in the Microsoft Decision “haultime” minutes 5.50 0.94 4.98 0.96
Tree approach. The DN displays the relation- “haulreturn” minutes 4.16 0.92 2.94 0.82
ships among attributes (input and predictable) “cycletime” minutes 2.88 0.84 2.39 0.77
from the decision tree model’s content and their “hauldist” meters 3.24 0.87 2.43 0.83
associated weights. The weights are related with