Research On Book Purchases System Based On Improved Genetic Algorithm And Neural Network

With the development of science and technology, the needs of human being to science and technology cultural life increase rapidly. The needs of human being for the life of science and technology culture life increase rapidly. Every library is expected to purchase books constantly in order to satisfy human needs, spread national excellent culture and use world advanced culture for reference.However, there are many disadvantages in the traditional book purchases methods, so it is essential and urgent to develop an intelligent book purchases method to improve the quality and efficiency of the book purchases. As a modern tool, the computer is undoubtedly employed in this process.In this dissertation, an intelligent book purchases scheme which combines the genetic algorithm with neural network is put forward and the corresponding prototype system is also built.The work done in this dissertation includes:[1] The existing book purchases methods are introduced and studied. Aiming the disadvantages in the process of traditional book purchases, we have discussed the feasibility about book purchases intelligence, which enlightens the further researches of the paper.[2] Aiming the disadvantage of traditional neural network training algorithms, slow convergence speed and easy to fall into local least spot, we have analyzed the common improved methods of neural network training. On the bases of studying and improving to genetic algorithm, we put forward genetic algorithm with effect factor (EFGA) which can be used to optimize neural network. In the processing of optimizing neural network with EFGA, the weighs, thresholds and structure of neural network can be adjusted dynamically so the neural network can converge to optimal spot rapidly. At last, a neural network with better generalization ability can be obtained.[3] Aiming at the disadvantages of traditional book purchases method, the paper demonstrates the feasibility of applying the neural network to the book purchases. Subsequently, a book purchases system model based on improved genetic algorithm and neural network is designed according to the characteristic of book purchases. The basic idea of this system model is training of neural network with the improved genetic algorithm to obtain the knowledge needed by the system and using the trained neural network to finish the book purchases reasoning work of the system.[4] At last, an intelligent book purchases tool—the book purchases prototype system based on the improved genetic algorithm and neural network is developed.