000			02310aab a2200217 4500
008			240527b20242024|||br||| |||| 00| 0 eng d
022			_a0889-3241
100			_aJu Dong Lee _9882107
100			_aThomas H.-K. Kang _9882108
245			_aUltimate Shear Strength Prediction for Slender Reinforced Concrete Beams without Transverse Reinforcement Using Machine Learning Approach
300			_a87-98 p.
520			_aA great deal of attention has been applied recently to machine learning (ML) algorithms to solve difficult engineering problems in the field of structural engineering. Using borrowed features of ML algorithms (implemented), a solution to one of the most troublesome problems in concrete structures—namely, shear—is proposed. The understanding of shear failure in reinforced concrete (RC) structures has led to numerous laboratory investigations and analytical studies over the last century. Due to decades of efforts afforded by researchers, significant experimental shear test results have been created and archived. This data provides an opportune environment to implement ML techniques and evaluate model efficiency and accuracy. The focus of this paper is on ML modeling of the shear-transfer mechanism for slender RC beams without transverse reinforcement. Test results for 1149 RC beams were incorporated in the ML analysis for training (80%) and testing (20%) purposes. Prior to the ML analysis, a correlation coefficient analysis was conducted to determine if given design parameters affected shear strength. When compared to the data used, code-based shear equations provided with large safety margins gave reasonable predictions. Exponential-based Gaussian process regression (GPR) ML models yielded comparable predictions. Of the 19 ML models employed, most were considered as an effective strength predictive tools. These ML model predictions were compared to each other and with design provision shear equations.
650			_aMachine Learning _9845
650			_aPrediction _9675736
650			_aRC Slender Beams _9882109
650			_aShear Database _9170979
650			_aShear Strength
773	0		_x08893241 _tACI Structural Journal _dFarmington Hills,MI, U.S.A : American Concrete Institute
856			_uDOI: 10.14359/51740246
942			_2ddc _n0 _cART _o14993 _pMr. Muhammad Rafique Al Haj Rajab Ali (Late)
999			_c815198 _d815198