Kernel-based models for prediction of cement compressive strength

Abstract This paper employs three different kernel-based models—support vector regression (SVR), relevance vector machine (RVM) and Gaussian process regression (GPR)—for the prediction of cement compressive strength. The input variables for the model are taken as C3S (%), SO3 (%), Alkali (%) and Blaine (cm2/g), while the output is 28-day cement compressive strength (N/mm2) of the cement. The hyperparameters of the SVR are obtained using two different metaheuristic optimization algorithms—particle swarm optimization (PSO) and symbiotic organism search (SOS). Trial-and-error-based approach is used for arriving at the hyperparameters of RVM and GPR. The compressive strength predicted using different kernel-based models is also compared with that obtained from ANN and fuzzy logic models reported in the literature. The performance of the different kernel-based models is benchmarked using six different error indices and residual analysis. The performance of the kernel-based models is found to be at par with ANN. The better generalization capability and excellent empirical performance of the kernel-based models overcome the disadvantages associated with ANN and provide a good tool for the prediction of the cement compressive strength.

[1]  Serhan Ozdemir,et al.  The use of GA-ANNs in the modelling of compressive strength of cement mortar , 2003 .

[2]  Theo Fett,et al.  Determination of Strength , 1999 .

[3]  Angel Irabien,et al.  Neural network prediction of unconfined compressive strength of coal fly ash–cement mixtures , 2003 .

[4]  Cheng-Lung Huang,et al.  A hybrid SOFM-SVR with a filter-based feature selection for stock market forecasting , 2009, Expert Syst. Appl..

[5]  Shervin Motamedi,et al.  Application of adaptive neuro-fuzzy technique to predict the unconfined compressive strength of PFA-sand-cement mixture , 2015 .

[6]  Rajib Maity,et al.  Potential of support vector regression for prediction of monthly streamflow using endogenous property , 2010 .

[7]  Dookie Kim,et al.  Application of support vector regression for the prediction of concrete strength , 2007 .

[8]  E. Badogiannis,et al.  A study on the parameters affecting the properties of Portland limestone cements , 1999 .

[9]  Gustavo Camps-Valls,et al.  Retrieval of Biophysical Parameters With Heteroscedastic Gaussian Processes , 2014, IEEE Geoscience and Remote Sensing Letters.

[10]  Vladimir Naumovich Vapni.  The Nature of Statistical Learning Theory , 1995 .

[11]  André Lecomte,et al.  Relation between cement composition and compressive strength of pure pastes , 2009 .

[12]  Min-Yuan Cheng,et al.  Symbiotic Organisms Search: A new metaheuristic optimization algorithm , 2014 .

[13]  Pijush Samui,et al.  Support vector regression based models to predict fracture characteristics of high strength and ultra high strength concrete beams , 2013 .

[14]  C. E. D. S. Tango.  An extrapolation method for compressive strength prediction of hydraulic cement products , 1998 .

[15]  Fa-Liang Gao,et al.  A new way of predicting cement strength—Fuzzy logic , 1997 .

[16]  Mahdi Hasanipanah,et al.  Prediction of air-overpressure caused by mine blasting using a new hybrid PSO–SVR model , 2016, Engineering with Computers.

[17]  Michael E. Tipping.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001, J. Mach. Learn. Res..

[18]  E F Vonndop.  THE INFLUENCE OF ALKALIS ON THE STRENGTH PROPERTIES OF PORTLAND CEMENT , 1979 .

[19]  Gustavo Camps-Valls,et al.  Retrieval of oceanic chlorophyll concentration with relevance vector machines , 2006 .

[20]  Chi-Jie Lu.  Hybridizing nonlinear independent component analysis and support vector regression with particle swarm optimization for stock index forecasting , 2012, Neural Computing and Applications.

[21]  Christopher K. I. Williams.  Prediction with Gaussian Processes: From Linear Regression to Linear Prediction and Beyond , 1999, Learning in Graphical Models.

[22]  Sunday O. Olatunji,et al.  Computational intelligence method of determining the energy band gap of doped ZnO semiconductor , 2016 .

[23]  S. Gunn.  Support Vector Machines for Classification and Regression , 1998 .

[24]  Vladimir Vapnik,et al.  An overview of statistical learning theory , 1999, IEEE Trans. Neural Networks.

[25]  Mauro Birattari,et al.  Swarm Intelligence , 2012, Lecture Notes in Computer Science.

[26]  Ketil Svinning,et al.  Prediction of potential compressive strength of Portland clinker from its mineralogy , 2010 .

[27]  Jian Zhang,et al.  Support vector regression for on-line health monitoring of large-scale structures , 2006 .

[28]  Gokmen Tayfur,et al.  FUZZY LOGIC MODEL FOR THE PREDICTION OF CEMENT COMPRESSIVE STRENGTH , 2004 .

[29]  Vagn Johansen,et al.  Particle Size Distribution and Rate of Strength Development of Portland Cement , 1989 .

[30]  Dario Rossi,et al.  Support vector regression for link load prediction , 2008, 2008 4th International Telecommunication Networking Workshop on QoS in Multiservice IP Networks.

[31]  Ta-Peng Chang,et al.  Estimation of exposed temperature for fire-damaged concrete using support vector machine , 2009 .

[32]  Shervin Motamedi,et al.  Estimating unconfined compressive strength of cockle shell-cement-sand mixtures using soft computing methodologies , 2015 .

[33]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[34]  Hadi Fattahi,et al.  Estimation of asphaltene precipitation from titration data: a hybrid support vector regression with harmony search , 2014, Neural Computing and Applications.

[35]  J. Chen,et al.  Superfine cement for improving packing density, rheology and strength of cement paste , 2012 .

[36]  I. Celik.  The effects of particle size distribution and surface area upon cement strength development , 2009 .