Open AccessMachine learning in economic planning: ensembles of algorithms
Author(s) -
Jun-Won An,
Alexey Mikhaylov,
Natalia Sokolinskaya
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1353/1/012126
Subject(s) - convergence (economics) , computer science , operator (biology) , algorithm , mode (computer interface) , rate of convergence , noise (video) , range (aeronautics) , mathematical optimization , machine learning , artificial intelligence , mathematics , key (lock) , engineering , biochemistry , chemistry , computer security , repressor , aerospace engineering , transcription factor , economics , image (mathematics) , gene , economic growth , operating system
The algorithm for machine learning of a transport type model is presented for the optimal distribution of tasks in safety critical systems operating in an automatic mode without operator participation. Safety critical systems in various application areas can operate in a wide range of modes - from pure manipulation by the operator prior to their autonomous execution of tasks as part of heterogeneous group. As it is shown by simulation studies of the adaptation algorithm generalized payment matrix of the transport model to the real preferences of the decision maker, even in conditions of significant noise measurements, the proposed algorithm for machine learning model leads to a fairly rapid convergence of estimates. Normalized error from the 15 th step does not exceed 10 percent. In this case, the rate of convergence of estimates is not an end in itself in the case of adaptive distribution of tasks in the group of algorithms; an important indicator is the convergence of solutions that exist above the convergence of estimates.