Regression differentiation and regression integration of finite series

Data analysis is a complex and multi-dimensional concept. It explained by both the objective complexity of data itself subjective nature expert analyzing them. Therefore, adequate formalization this requires completely new apparatus, on one hand, capable overcoming (irregularity inaccuracy), and, other vague judgment. The development Discrete Mathematical Analysis (DMA) an important step in direction. DMA highly oriented occupies intermediate position between strict mathematical methods (statistical analysis, TFA, etc.) soft combinatorial (simulation modeling, neural networks, etc.). This paper proposes constructions regression derivatives integrals for discrete time series defined general irregular grid. An role their study played recently created authors projection method solving systems linear algebraic equations described at end paper. obtained have hierarchical character spirit wavelet fractal analysis. results work define another direction further studies, namely, propagation differentiation integration finite mathematics under scenarios classical one.