A Constant Binomial Coefficient Difference Equation Based FIR Predictor System for Signal Processing and Data Mining

Prediction/estimation of future probable outcomes of a time-dependent signal/process has applications in signal processing, automatic control systems, artificial intelligence (AI), machine learning algorithms, and planning and development, etc. In this chapter, a novel simple and new prediction algorithm based on FIR system using constant binomial coefficient difference equation approach has been proposed and its characteristic features have been discussed. The main advantage of this algorithm is that it can be realized using cascade-type FIR filter structure. While estimating the future sample values, the system takes into account most recent observations and gives more weightage to these in comparison to the distant past observations. Hence, these systems inherently possess AI features.