Artificial Neural Network Models for Predicting California Bearing Ratio of Lateritic Soil Admixed with Reinforce and Rice Husk Ash

California bearing ratio (CBR) is an indispensable parameter in the design of road pavement, repeated carrying out of this test has been chiefly monotonous and time wasting, also the use of cement as stabilizer has also been increasingly expensive, hence, the need for admixing with agrowaste ash such as rice husk ash (RHA). This research is carried out for the prediction of the CBR of lateritic soil admixed with cement and RHA by means of an artificial neural network (ANN). Six parameters are selected as input variables to obtain results that are accurate and precise. The six input variables are cement, RHA, liquid limit, plasticity index, maximum dry density and optimum moisture content, while CBR Unsoaked and CBR Soaked are the output variables. The study consists of a database of 1288 samples obtained from laboratory experiments which were subdivided into 70% for training, 15% for testing, and 15% for validation. The training operation is performed by a multilayer perceptron-back propagation algorithm. The network topology is achieved after fixing a number of hidden neurons. Thereafter, statistical indices are used in evaluating the performance of the ANN model. It is established that this model is appropriate for accurate prediction of CBR results.