Method for predicting projectile speed based on machine learning

The invention relates to a learning method adopting a neural network, in particular to a method for predicting the speed of a projectile based on machine learning, which is used for solving the defect that the speed of the projectile is difficult to predict and calculate in a traditional mode. According to the method for predicting the projectile speed based on machine learning, firstly, a clustering algorithm is adopted to detect and delete abnormal data samples in a data sample set of projectile launching, then a plurality of projectile speed prediction models are established according to correlation among variables of the data sample set, and cross validation and model evaluation are conducted on the multiple projectile speed prediction models. And finally, establishing a layered integration model by using a plurality of projectile speed prediction models to generate an integrated prediction result. 本发明涉及采用神经网络的学习方法，具体涉及一种基于机器学习预测弹丸速度的方法，用于解决传统方式难以对弹丸的速度进行预测计算的不足之处。该基于机器学习预测弹丸速度的方法首先采用聚类算法检测并删除弹丸发射的数据样本集中的