Scholarlink

In this paper we design a recurrent neural network for the prediction of time series, In this case, we consider the Bitcoin time series. The objective is to find the best architecture and offer a good prediction error, the fuzzy integration is carried out with type-1, type-2, and type-3 fuzzy systems and a comparison between them is presented. The simulation results of this method produce good prediction errors since recurrent neural networks are effective techniques for data series.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparison of Type-1, Type-2 and Type-3 Fuzzy Integrators for Ensemble Neural Networks Applied to Bitcoin Prediction

  • 章节

  • 发表日期：

    2025年

查看全文
  • Martha Pulido,
  • Patricia Melin

摘要

In this paper we design a recurrent neural network for the prediction of time series, In this case, we consider the Bitcoin time series. The objective is to find the best architecture and offer a good prediction error, the fuzzy integration is carried out with type-1, type-2, and type-3 fuzzy systems and a comparison between them is presented. The simulation results of this method produce good prediction errors since recurrent neural networks are effective techniques for data series.