Mathematical analysis, advanced algebra, probability statistics. The main problem in mathematical modeling is nothing more than these blocks in knowledge points: 1, the problem of finding the maximum value of multivariate variables, which can finally be transformed into Lagrange multiplier method; 2. High-dimensional linear programming and linear regression problems are solved by matrix multiplication of linear algebra; 3. It may be necessary to use the relevant knowledge of stochastic process, apply the theorem of large numbers and Monte Carlo algorithm, and use probability statistics as a tool to solve it.