prediction of subgrade reaction modulus of clayey soil using group method of data handling

Settlement-based design of foundations, using subgrade reaction modulus (ks), is an important technique in geotechnical engineering. plate load test (plt) is one of the most commonly performed tests to directly obtain ks of soil. as the determination of ks by plt -especially in greater depths- is both time-consuming and costly, it is necessary to develop models that can easily handle measurable characteristics. the suitability of the group method of data handling (gmdh) polynomial neural network (nn) to estimate ks of clayey soils was investigated in the present research. in order to develop the gmdh models, 123 data sets from qazvin, iran, were applied. predictability of the derived equations was compared with those of other available equations for clayey soil. the results demonstrated that an improvement in predicting ks was achieved. sensitivity analysis of the best gmdh-based equation showed that the liquid limit (ll) of soil was the most in uential parameter on the proposed gmdh model in predicting ks of clayey soil.