The Assessment Of The Performance Of The Investment Funds Based On Neural Networks

As the result of financial innovation of security market, the security investment fund has a history of more than half a century abroad. The security investment fund plays an important role in the world economy. It is not only the voltage stabilizer of the capital market, but also a helping hand device of the international economy. With the globalization process of the security investment fund, the security investment fund of our country has been developed quickly. The theory and principles of investment of security investment fund have become attractive to more and more investors. In order to protect investors' investment interests, and meanwhile provide the basis for administrator's investment decision of the fund, it has more important realistic meanings to evaluate the achievement of the security investment fund. The so-called investment funds are the set of benefits-sharing, risks-sharing system. It makes the investors' fund together, and transfers it to the fund mandator to deposit through issuing the fund securities. It is managed by the fund caretaker. The fund is invested in financial instruments such as the stock, bond, and many others in order to obtain the investment return and capital appreciation. There is a popular saying, "everybody pools money and buys the securities, enjoy benefits together, stick together through thick and thin". It is different to some extent to invest fund in different countries or the regional appellation. U.S.A. is called "common fund", Britain and Hong Kong are called "unit's trust fund", Japan and Taiwan are called "the trust fund of investment securities".　This present works are chosen from 15 of 48 pieces of close-ended fund of December 28, 2001. Among them, the net value of total assets of 48 funds among them is 67 billion yuan. The text sets about from the current situation of study on fund, and takes 15 funds as the samples. They are, Kaiyuan fund, Longyu fund, Pudeng fund, Tongyi fund, Jinghong fund, Yuhua fund etc. The 15 close-ended fund account for 31.3%. The total fund assets net value accounts for 55.1% of the assets of net value. The sample fund can basically reflect the overall picture of the Chinese security investment fund. This thesis introduces the artificial neural network mainly ( Artificial Neural Network, abbreviate by ANN as (Error Back Propagation Algorithm, abbreviated as BP law) and join OPA (optimum split method ( optimal partition algorithm, RBF (because of radial base function (Radial Basis function-RBF)) on OPA)) algorithm turn on to fund such 15 pieces of close-ended fund as yuan ,etc. to predict. BP algorithm is a kind of study algorithm with a tutor. It is a kind of study...