Investment risk evaluation of hi-tech projects based on projection pursuit model

The hi-tech projects´ investment is characterized with high risk and huge profit, so a scientific and accurate risk evaluation model is of importance to projects´ investment. Based on the analysis of the indicators influencing the investment risk of hi-tech projects, some typical indicators were selected. Firstly, according to the quantitative assessment results from the experts, some samples were used to establish the projection pursuit model for comprehensive evaluating the investment risk of the hi-tech projects. Then, the model without samples was constructed to avoid the impact of training samples. The optimal projection vector was optimized by genetic algorithm, so the above problem was transformed into one dimension problem. Compared with the conventional assessment methods such as AHP (Analytic Hierarchy Processing) and ANN (artificial neural net), the model can effectively overcome the faults of the subjective feature from the AHP and unsatisfied generalization ability from ANN. Additionally, under no training samples, the model can also sort well, so the model has wide applicability.