Fuzzy Time Series

Fuzzy time series forecasting (FTSF) aims to predict future values in time series data using fuzzy logic, offering a flexible approach to handle uncertainty and non-linearity. Current research focuses on improving FTSF accuracy by addressing limitations of traditional models through techniques like incorporating particle swarm optimization, convolutional neural networks (including differential and semi-asymmetric variations), and dimensionality reduction methods such as embeddings. These advancements enhance the ability to capture long-term dependencies and handle high-dimensional data, leading to improved forecasting performance in diverse applications, such as energy consumption prediction in smart grids.