The invention discloses a heterogeneous domain adaptation method based on divergence calculation. The method comprises the following steps of: preprocessing data; respectively learning and mapping, mapping a source domain and a target domain to the same shared subspace, performing distribution matching by using a maximum mean difference (MMD) distance, reducing distribution difference, and performing nucleation processing to adapt to a nonlinear problem; following the structural consistency, guaranting that similar data are kept similar after mapping; introducing a regular item for feature selection; the separability among different types of data is further improved through divergence calculation, and the algorithm efficiency is improved; integrated optimization. According to the method, atransfer learning thought is adopted, and mapping and feature conversion are carried out on heterogeneous source domain data and target domain data respectively to obtain subspaces with similar features and distribution. Divergence is introduced into heterogeneous application, data separability is further enhanced by means of divergence characteristics, classification of heterogeneous data with different characteristics and different distributions is enhanced, and the learning effect of a target task is improved.