Resumen

RIABANI MERCADO, Franklin; GARCIA FERNANDEZ, Willman  y  HERRERA ACEBEY, Johnny A.. Artificial intelligence system for early prediction of weather frost. RevActaNova. [online]. 2016, vol.7, n.4, pp.483-495. ISSN 1683-0789.

Frost is an important weather factor for hydrology, climatology and agriculture. This study proposes to create a learning machine end (ELM) using the method proposed by Huang on a layer neural network monolayer as a basis for early frost weather prediction algorithm. The study was developed in the Valle Alto of Cochabamba-Bolivia, data were collected at 6 meteorological stations reaching a total of 178450 measurements, for training the neural network. For verification, meteorological data were used stations from Arque and Tiquipaya 2016. With confidence margins above 90% they were used. As research results showed that a neural network trained with the proposed Huang algorithm, it is a good predictor of weather frost from both, confidence levels and times response.

Palabras clave : Computational Intelligence; Artificial Intelligence; Machine Leaming; SciPy; Raspberry Pi; intelligent Agriculture; Frosts.

        · resumen en Español     · texto en Español     · Español ( pdf )