In the one hand, Prediction and knowing about future for making management decisions and providing strategic politics is very important. on other the hand, surviving in competitive marketing in the variety industries such as clothing industry is necessary. Due to this reason, quality prediction of production is one of the most important things which affects surviving competitive marketing and increasing this and it can be important for managers of clothing industries. Due to the importance of predicting the competitive environment governing the textile industry, various models have been proposed to improve the accuracy of predictions. In this regard, artificial neural networks are precision tools for a wide range of issues, which require much data to achieve accurate results, limiting their use. However, providing the required data is fundamentally costly and time-consuming. Hence, the use of methods that can provide prediction by a small number of data can be more appropriate and efficient in such markets. Computational intelligence and soft computing methods are the most accurate and most widely used methods for modeling the complexity and uncertainty in the data. In this dissertation, in order to avoid high modeling costs, avoiding the problem of excessive fitting of neural networks in modeling linear patterns and also eliminating the limitations of nonlinear patterns modeling in statistical classical methods, a hybrid model based on multivariate linear regression models, neural networks Multilayer perceptron, logic, and fuzzy numbers, simultaneously taking advantage of the combination of series and parallel structure, to predict seam clothing quality. The results obtained from using the proposed method indicate a higher performance of this method in relation to its constituent models as well as other combination methods mentioned in this thesis. Therefore, it can be claimed that the proposed hybrid model can be used as a useful tool for predicting seam clothing quality.