One of the main objectives of chemical stabilisation is to increase the compressive strength of soils. A wide range of parameters affect the strength improvement in cementitious stabilisation with chemicals. Accordingly, it is difficult to determine some kinds of functional relationships in strength improvement which make the precision of strength prediction to be satisfying. The purpose of the present study is to use two computational intelligence techniques namely, multilayer perceptron (MLP) and linear genetic programming (LGP), in order to develop the mathematical models to be capable of predicting the unconfined compressive strength. Subsequently, a comparison between these methods was performed in terms of prediction performance. Properties of natural soil such as textural properties, plasticity and linear shrinkage, stabiliser quantities and types (cement, lime, asphalt), for a wide range of soil types were used in order to generate the mathematical models to be able to predict the compressive strength as a quality of stabilised soil. A comprehensive set of data including 219 previously published stabilised unconfined compressive strength experimental determinations were utilised to develop the models.@en

Different parameters influence the compressive strength and workability properties of high performance concrete (HPC) mixes. Accordingly, an extensive understanding of relation between these parameters and properties of the resulting matrix is required for developing a standard mix design procedure for HPC mix. The complex behaviour of strength and workability improvement and a need to avoid trying several mix proportions to generate a successful mix suggest the necessity to develop comprehensive mathematical models to be able to evaluate the performance characteristics of HPC mixes with high accuracy. Therefore, in this paper, linear genetic programming (LGP) is utilised for the first time in the literature, to develop mathematical models to be able to predict the strength and slump flow of HPC mixes from the influencing parameters. Subsequently, the LGP based prediction results are compared with the results of proposed multilayer perceptron (MLP) in terms of prediction performance. Sand cement ratio, coarse aggregate cement ratio, water cement ratio, percentage of silica fume and percentage of superplasticiser are used as the input variables to the models to predict the strength and slump flow of HPC mixes. A reliable database was obtained from the previously published literature in order to develop the models.@en