TITLE:
M5 Model Tree to Predict Temporal Evolution of Clear-Water Abutment Scour
AUTHORS:
R. Biabani, M. Meftah Halaghi, Kh. Ghorbani
KEYWORDS:
Abutments, Scour Depth, M5 Model Tree, Genetic Programming Model (GP)
JOURNAL NAME:
Open Journal of Geology,
Vol.6 No.9,
September
9,
2016
ABSTRACT: Scour
is a natural phenomenon that is created by the rivers streams or the flood
which brings about transferring or eroding of bed materials. To have accurate
and safe erosion control structures design, maximum scour depth in downstream
of the structures gains specific
significance. In the current study, M5 model tree as remedy data mining
approaches is suggested to estimate the scour depth around the abutments. To do
this, Kayaturk laboratory data (2005), with different hydraulic conditions, are
used. Then, the results of M5 model were also compared with genetic programming
(GP) and pervious empirical results to investigate the applicability, ability,
and accuracy of these procedures. To
examine the accuracy of the results yielded from the M5 and GP procedures, two
performance indicators (determination coefficient (R2) and root mean square
error (RMSE)) were used. The comparison test of results clearly shows that the
implementation of M5 technique sounds satisfactory regarding the performance indicators (R2 = 0.944 and
RMSE = 0.126)
with less deviation from the numerical values. In addition, M5 tree model, by
presenting relationships based on liner regression, has good capability to
estimate the depth of scour abutment for engineers in practical terms.