Theoretical and empirical research on investment behavior strategy based on fuzzy probability

Aiming at of "social" learning characteristics and relatively strong local searching ability of PSO, excellent performance of GSA on exploration ability and characteristics of skipping from local limit of CS for ergodicity and randomness, the Paper will integrate the three kinds of algorithms into a kind of new hybrid meta-heuristic algorithm (HMA) and apply it to solution of investment combination optimization under cardinality constrain and test result of standard test function shows that HMA has better optimization performance, higher searching precision and faster convergence speed compared with PSO and GSA; at the same time, empirical analysis also verifies that it is effective to solve CCPS by the algorithm and it will achieve better result compared with GA and PSO and will acquire solution with higher quality. Follow-up work of the Paper is to solve other non-smooth financial optimization problems by adopting the algorithm to acquire better and wider application.