Predicting events by analyzing the results of the work of predictive models

A Prangishvili, Z Gasitashvili, M Pkhovelishvili… - … Conference on Pattern …, 2021 - Springer
Building much more effective new hybrid models from prediction models is discussed. The
algorithm for selection of model pairs, triplets, etc. and the advantage of obtained model over …

Theory of universal approach to improve predictive models using parallel data and application examples

A Prangishvili, Z Gasitashvili, M Pkhovelishvili… - AIP Conference …, 2023 - pubs.aip.org
In this article we discuss the technique of creating a new, much better predictive model from
existing predictive models by using parallel data. There is given theoretical proof of its …

Peculiarities of the Hybrid Model Built Using Parallel Data

M Phkhovelishvili, N Archvadze… - 2022 7th International …, 2022 - ieeexplore.ieee.org
A hybrid model built using parallel data combines the prerequisites of two or more models in
its predecessors, it becomes necessary to reduce the number of these predecessors. For …

[PDF][PDF] New algorithms for improvement of prediction models using data parallelism

Z Gasitashvili, M Phkhovelishvili, N Archvadze - … International Conference on …, 2021 - csit.am
The paper reviews the issues of improvement of existing prediction models, for which
parallel data is used. These are the data, which simultaneously impact prediction of some …

Modern forecasting models in economy

N Archvadze, M Pkhovelisvili - X International Conference of the …, 2019 - researchgate.net
A new approach for business forecasting is discussed in the article. This means usage of
parallel data paradigm of programming. Parallel data are different kind of former data, which …

[引用][C] New algorithm for building effective model from prediction models using parallel data