Urban activity Logit model estimation method oriented to oversized samples and multiple options

According to the urban activity Logit model estimation method oriented to the ultra-large samples and the ultra-multiple options, sample data of urban activities are counted into a corresponding matrix form, the input data size of the model is greatly reduced, then parameters of the corresponding Logit model are estimated through a matrix calculation mode based on mathematical derivation, and under the condition that the low memory overhead is guaranteed, the estimation efficiency is improved. The urban activity Logit model estimation method oriented to oversized samples and multiple options is realized. The method not only greatly reduces the problem of information repetition in the input data, but also solves the problems of insufficient memory and poor computing power of a common computer for big data. And meanwhile, the calculation speed can be greatly improved, and basically all computers can quickly complete the task, so that the method has relatively high applicability. 本发明提供一种面向超大样本和超多选项的城市活动Logit模型估计