An R implementation of a Recurrent Neural Network Trained by Extended Kalman Filter

Journal Title: Revista Romana de Statistica - Year 2016, Vol 64, Issue 2

Abstract

Nowadays there are several techniques used for forecasting with different performances and accuracies. One of the most performant techniques for time series prediction is neural networks. The accuracy of the predictions greatly depends on the network architecture and training method. In this paper we describe an R implementation of a recurrent neural network trained by the Extended Kalman Filter. For the implementation of the network we used the Matrix package that allows efficient vector-matrix and matrix-matrix operations. We tested the performance of our R implementation comparing it with a pure C++ implementation and we showed that R can achieve about 75% of the C++ programs. Considering the other advantages of R, our results recommend R as a serious alternative to classical programming languages for high performance implementations of neural networks.

Authors and Affiliations

Bogdan Oancea, Tudorel Andrei, Raluca Mariana Dragoescu

Keywords

Related Articles

Evoluţia preţului unei acţiuni bazată pe analiza activităţii online a investitorilor

Modelarea influenţei factorilor proveniţi din economia comportamentală, reprezentând de cele mai multe ori variabile calitative, impune abordări noi pentru a cuantificarea influenţei respective. Se prezintă o metodă de a...

Disparitate şi specificitate a preţurilor la consumator

Statistica este ştiinţa de cunoaştere cu cea mai mare sferă de cuprindere de aplicare şi utilizare a metodelor sale de culegere, prelucrare şi prezentare de date şi informaţii din diferite ramuri de activitate: industrie...

A NEW METHOD OF STATISTICAL ANALYSIS OF MARKETS’ CONCENTRATION OR DIVERSIFICATION

This paper tries to clarify, from a statistical point of view, an important notion like “market with excessive concentration” (defining the relative limit of a concentration coefficient Gini - Struck, in the general co...

Some Consideration in Defining Management Modernization Strategies

We can appreciate that competitive strategy is “faster, better, more” and in agribusiness field have an strategic importance. To maximize the value chain components requires good connection and ensure their coordination,...

THE EVOLUTION OF ROMANIA’S GROSS DOMESTIC

This paper describes the most important aspects regarding the evolution of the Gross Domestic Product of Romania. The main topics approached by the author are the modification factors on categories of resources, the evolu...

Download PDF file
  • EP ID EP138936
  • DOI -
  • Views 157
  • Downloads 0

How To Cite

Bogdan Oancea, Tudorel Andrei, Raluca Mariana Dragoescu (2016). An R implementation of a Recurrent Neural Network Trained by Extended Kalman Filter. Revista Romana de Statistica, 64(2), 125-133. https://europub.co.uk/articles/-A-138936