IP3 2023 – Artificial Intelligence & Machine Learning 1 (LOT 14911)

This lot is generally related to pre-processing input data to a non-linear predictive system model based on a support vector machine. Disclosed is a data pre-processor for pre-processing input data for a support vector machine, having outlier values consisting of an input buffer which receives and stores the input data. The system has a data filter for detection and removal of multiple outlier values and generates corrected input data having corrected run-time data. Also disclosed is an output device for outputting the corrected input data which has the input data to the support vector machine. The support vector machine has a non-linear model with a set of model parameters for defining a representation of a system using a training model. The support vector machine is operated using the received and corrected run-time data and generate run-time output data. The technology may be implemented in eCommerce systems, financial markets, data mining systems, etc.