Research on Application of a New Algorithm Based on Multi-source Data in Tourism Flow Forecasting

Currently(1), there are many prediction algorithms in terms of passenger flow volume of scenic spots. In this paper, it studies and analyzes the connection between multi-source search data and passenger flow volume of scenic spots, proposes application of a new algorithm integrating non-linear programming, genetic algorithm and support vector regression in the passenger flow volume prediction, fully takes advantages of non-linear function and genetic algorithm, quickly calculates the optimal parameter of kernel function on the basis of not adding calculated amount, and then employs support vector regression algorithm to calculates prediction result, and finally uses the actual travel data collected to verity the accuracy of this prediction model.