To estimate the inPUT-Output mapping with inputs x
％ and outputs y generated by the following nonlinear,
％ nonstationary state space model:
％ x(t+1) = 0.5x(t) + [25x(t)]/[(1+x(t))^(2)]
％ + 8cos(1.2t) + process noise
％ y(t) = x(t)^(2) / 20 + 6 squareWave(0.05(t-1)) + 3
％ + time varying measurement noise
％ using a multi-layer perceptron (MLP) and both the EKF and
％ the hybrid importance-samping resampling (SIR) algorithm.
Batch version of the back-propagation algorithm.
％ Given a set of corresponding inPUT-Output pairs and an initial network
％ [W1,W2,critvec,iter]=batbp(NetDef,W1,W2,PHI,Y,trparms) trains the
％ network with backpropagation.
％
％ The activation functions must be either linear or tanh. The network
％ architecture is defined by the matrix NetDef consisting of two
％ rows. The first row specifies the hidden layer while the second
％ specifies the output layer.
％
All of Java s Input/Output (I/O) facilities are based on streams, which provide simple ways to read and write data of different types. Java provides many different kinds of streams, each with its own application. The universe of streams is divided into four large categories: input streams and output streams, for reading and writing binary data and readers and writers, for reading and writing textual (character) data. You re almost certainly familiar with the basic kinds of streams--but did you know that there s a CipherInputStream for reading encrypted data? And a ZipOutputStream for automatically compressing data? Do you know how to use buffered streams effectively to make your I/O operations more efficient? Java I/O, 2nd Edition has been updated for Java 5.0 APIs and tells you all you ever need to know about streams--and probably more.
