FPGA realization of backpropagation for stock market prediction

In this paper, we present the realization of backpropagation on Altera FLEX10K FPGA device for stock market prediction utilizing the parallelism that exists in the neural network architecture. This approach provides an increased speed of convergence of the network and accuracy for the stock market forecast. The stock market prediction neural network architecture comprises of three layers, input layer, hidden layer and output layer. There are three neurons in the input layer, two neurons in the hidden layer and one neuron in the output layer. Sigmoid transfer function is used for hidden layer and output layer neuron. Neuron for each of the backpropagation layer is modeled individually using behavioral VHDL. The layers are then connected using structural VHDL. This is followed by the timing analysis and circuit synthesis for the validation, functionality and performance of the designated circuit which supports the practicality, advantages and effectiveness of the proposed hardware realization for the applications.

[1]  A. Garliauskas,et al.  Neural network chaos and computational algorithms of forecast in finance , 1999, IEEE SMC'99 Conference Proceedings. 1999 IEEE International Conference on Systems, Man, and Cybernetics (Cat. No.99CH37028).

[2]  C. E. Cummings.  Verilog simulation of Xilinx designs , 1994, International Verilog HDL Conference.

[3]  Nadipuram R. Prasad,et al.  Neural Networks and Fuzzy Logic , 1998 .

[4]  Juan M. Corchado,et al.  Unsupervised learning for financial forecasting , 1998, Proceedings of the IEEE/IAFE/INFORMS 1998 Conference on Computational Intelligence for Financial Engineering (CIFEr) (Cat. No.98TH8367).

[5]  Clarence N. W. Tan,et al.  A study of the parameters of a backpropagation stock price prediction model , 1993, Proceedings 1993 The First New Zealand International Two-Stream Conference on Artificial Neural Networks and Expert Systems.

[6]  Ralph Neuneier,et al.  Multi-Agent FX-Market Modeling by Neural Networks , 2002 .

[7]  R. Kane,et al.  Financial forecasting and rules extraction from trained networks , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[8]  K. Lagemann,et al.  NeNEB-an application adjustable single chip neural network processor for mobile real time image processing , 1996, Proceedings of International Workshop on Neural Networks for Identification, Control, Robotics and Signal/Image Processing.

[9]  B.K. Fawcett.  Tools to speed FPGA development , 1994, IEEE Spectrum.

[10]  C. Dagli,et al.  Simple ensemble-averaging model based on generalized regression neural network in financial forecasting problems , 2000, Proceedings of the IEEE 2000 Adaptive Systems for Signal Processing, Communications, and Control Symposium (Cat. No.00EX373).

[11]  Ralph Neuneier,et al.  Multi-agent FX-Market Modeling Based on Cognitive Systems , 2001, ICANN.

[12]  Alexandre Schmid,et al.  Hardware realization of a Hamming neural network with on-chip learning , 1998, ISCAS '98. Proceedings of the 1998 IEEE International Symposium on Circuits and Systems (Cat. No.98CH36187).