The Joint Use of Artificial Intelligence Techniques for Diagnostication and Prediction

Abstract

The paper presents some aspects regarding thejoint use of artificial intelligence techniques for the activityevolution diagnostication and prediction by means of a set ofindexes. Starting from the indexes set a measure on thepatterns set is defined, measure representing a scalar valuethat characterizes the activity analyzed at each time moment.A pattern is defined by the values of the indexes set at a giventime. Over the classes set obtained by means of theclassification and recognition techniques is defined a relationthat allows the representation of the evolution from negativeevolution toward positive evolution. For the diagnosticationand prediction the following tools are used here: regressionalmodels, pattern recognition and multilayer perceptron. Thedata set used in experiments describes the evolution of theBucharest Stock Exchange (BSE). The paper presents:REFORME software written by the authors and theexperiments carried out in order to analyze the activity ofBSE.

Keywords

• Pattern recognition
• Neural networks
• Multilayer perceptron
• Regression
• Diagnostication
• Prediction