System and method for generating a probability value for an event

1. A non-transitory computer-readable medium that stores a program that causes a processor to perform functions to generate a probability value for a seizure based on an EEG by executing the following steps:

receiving an EEG recording at a first processor, the EEG recording generated from a plurality of EEG signals generated from an EEG machine comprising a plurality of electrodes, at least one amplifier, and a second processor, wherein the EEG recording comprises a plurality of channels corresponding to the plurality of electrodes, each of the plurality of channels comprising a plurality of epochs;

submitting the EEG recording to a multilayer perceptron operating on a first processor to generate a plurality of raw score epochs, wherein the multilayer perceptron comprises six inputs, three hidden nodes and a single output;

sorting at the first processor the plurality of raw score epochs by similar values;

calibrating at the first processor a first half of the plurality of raw score epochs with similar values to generate a probability value that a seizure will occur;

validating at the first processor the probability value that a seizure has occurred with actual data;

validating at the first processor a second half of the plurality of raw score epochs with similar values at the multilayer perceptron to determine if the probability value is correct; and,

generating at the first processor a graph of the probability value versus time on a graphical display.

2. A system for generating a probability value for a seizure based on an EEG, the system comprising:

an EEG machine comprising a plurality of electrodes and at least one amplifier;

a processor connected to the EEG machine to receive a plurality of EEG signals from the EEG machine; and

a graphical display connected to the processor;

wherein the EEG recording comprises a plurality of channels corresponding to the plurality of electrodes, each of the plurality of channels comprising a plurality of epochs;

wherein the EEG recording is submitted to a multilayer perceptron at the processor to generate a plurality of raw score epochs, wherein the multilayer perceptron consist essentially of six inputs, three hidden nodes and a single output;

wherein the processor is configured to sort the plurality of raw score epochs by similar values;

wherein the processor is configured to calibrate a first half of the plurality of raw score epochs with similar values to generate a probability value that a seizure will occur;

wherein the processor is configured to validate the probability value that a seizure has occurred with actual data of a seizure;

wherein the processor is configured to validate a second half of the plurality of raw score epochs with similar values the multilayer perceptron to determine if the probability value is correct; and,

wherein the processor is configured to generate a graph of the probability value versus time on the graphical display.

3. A non-transitory computer-readable medium that stores a program that causes a processor to perform functions to determine a probability value for fraudulent credit card transactions by executing the following steps:

receiving a plurality of credit card transactions at a server, the server comprising a plurality of peripheral interfaces, a processor, a memory, a storage, a communication module and an output module;

submitting to the server the plurality of credit card transactions to a multilayer perceptron to generate a plurality of raw score value for each of the credit card transactions, wherein the multilayer perceptron comprises six inputs, three hidden nodes and a single output;

sorting at the server the plurality of raw score values by similar values;

calibrating a first half of the plurality of raw score values with similar values to generate a probability value that a credit card transaction was fraudulent;

validating at the server the probability value that a fraudulent credit card transaction has occurred with actual data;

validating at the server a second half of the plurality of raw score values with similar values at the multilayer perceptron to determine if the probability value is correct; and,

generating at the server a graph of the probability value versus time on a graphical display.