Prediction of Pore Water-Pressure Using Multiple Linear Regression (MLR)

Knowledge of pore water pressure responses to rainfall is requires in any project
related to slopes stability analysis. Multiple Linear Regression (MLR) is an extension
of linear regression and can be used to represent the relationship between dependent
and independent variables. Multiple Linear Regression (MLR) was employed to
develop the MLR model equation for prediction of pore water pressure responses to
rainfall. For this purpose, the time series of pore water pressure and rainfall data
were obtained and used to develop the MLR model. Total of 1416 data were
collected through a field work of a slope in Universiti Teknologi PETRONAS. Data
was divided into two groups as 990 for training and 426 for testing of pore water
pressure. The input data was analyzed using linear regression and the performance of
the MLR model was evaluated by different statistical measures such as the analysis
of variance (ANOVA), coefficient of determination (R2), root mean square error
(RMSE) and the coefficient of efficiency (CE). Prediction results showed that MLR
model equation has a better performance during training since testing has produced a
high value of coefficient of determination,