The results of the review and the factors used to predict a total of 14 factors, including economic fluctuations. Labor shortage For the completion of construction Features of Material / Equipment Delays before starting projects The information necessary for construction Preparation Time The delay in approval The delay on the part of the owner. Job changes Construction, not the design. Contractors abandon the protest, complaint, conflict and a breakdown in communications. And other expenses incurred by the project. This led to the emergence of all the factors used to predict the effect on neural network learning, spreading to and function and transfer function results sigmoid to determine the percentage of statements. Reserve estimated the value of the contract. The result is a neural network with the best performance in predicting the outcome can be predicted with a contract value in the range of 0-500 million baht, 500-1,000 million baht, 1000-1500, and more than 1,500 million baht will be. Error value equal to 19.52%, 69.88%, 1.90% and 10.59%, respectively.
การแปล กรุณารอสักครู่..
