BetOracle takes decision whether or not the bet should be placed.
This decision is based on a set of binary inputs, e.g. current price,
totalToBack, totalToLay, etc. Each input value has to be normalised as a
value between 0(inclusive) and 1(inclusive). BetOracle uses a feed forward
neural and resilient propagation training for learning process. It returns
one integer value: 1 or 0. 1 - back bet should be placed, 0 lay bet should
be placed.
Configuration:
- inputs: Number of binary inputs that are provided to the BetOracle for
bet placement decision.ive)
Functions:
/** Calculates bet placement decision. Either back or lay bet should be placed.
*
* @param inputs Values of the binary inputs that the bet placement
decision is taken for.
* @return 1 or 0. 1 - back bet should be placed, 0 lay bet should be placed.
*/
def compute(inputs: Array[Double]):Double
**************************************************************
/** Train the BetOracle what was the correct bet placement decision for the
given set of inputs.
*
* @param inputs Values of binary inputs that the bet placement decision
was taken for.
* @param output Correct bet placement decision that should be taken for
that bet. It should take values 1 or 0.
* 1 - placing back bet was the correct decision, 0 - placing lay
bet was the correct decision.
*/
def train(inputs: Array[Double], output: Double)
***************************************************************
/**
* @return Get the current error percent from the training. Between 0 and 1.
*/
def getError():Double
Original issue reported on code.google.com by daniel.k...@gmail.com on 24 Mar 2010 at 1:33
Original issue reported on code.google.com by
daniel.k...@gmail.com
on 24 Mar 2010 at 1:33