2234 Figure 11 – Integrated Execution The methodology used to calculate the TPH improvement is based on a structured, data-driven analytical framework that integrates clustering, statistical modeling, and incremental impact estimation First, operational data with 15-minute granularity is segmented through deep learning–based clustering to group periods with similar mineralogical and operating conditions. This ensures that project comparisons are performed under comparable ore feed characteristics, reducing bias caused by variability in geology or process conditions. Figure 12 – Data clustering Once the events and project implementation dates are identified in the operational timeline, a multivariable linear regression model is developed to evaluate statistical significance and quantify the effect of each factor on SAG mill tonnage
RkJQdWJsaXNoZXIy MTM0Mzk2