Machine learning technology is widely used in the field of enterprise intangible value assessment due to its advantages in processing complex data and discovering linear relationships. This paper designs a B-P neural network model based on machine learning technology and compares it with the cost method, market method, income method, and option pricing B-S model for enterprise intangible asset value assessment. The performance of this paper’s model for predicting intangible assets is evaluated through enterprise transaction data collection and processing. In the training iteration of 40–80 rounds, this paper’s model loss, RMSE, accuracy, and recall successively converge to 0.08, 0.17, 0.91, and 0.96, and the relative error for the prediction of the value of enterprise intangible assets is low, which has a high performance in intangible value assessment. Additionally, this paper’s model computes the intangible asset value weights for various enterprises, and the results of expert judgments are essentially consistent. For instance, when analyzing human capital experts, the model calculates weights of 27.93% and 29.43%, respectively. This paper provides a scientific and accurate machine learning technology for enterprise intangible asset value assessment.