Abstract:Optimal decision-making in social settings is often based on forecasts from time series (TS) data. Recently, several approaches using deep neural networks (DNNs) such as recurrent neural networks (RNNs) have been introduced for TS forecasting and have shown promising results. However, the applicability of these approaches is being questioned for TS settings where there is a lack of quality training data and where the TS to forecast exhibit complex behaviors. Examples of such settings include financial TS forecasting, where producing accurate and consistent long-term forecasts is notoriously difficult. In this work, we investigate whether DNN-based models can be used to forecast these TS conjointly by learning a joint representation of the series instead of computing the forecast from the raw time-series representations. To this end, we make use of the dynamic factor graph (DFG) to build a multivariate autoregressive model. We investigate a common limitation of RNNs that rely on the DFG framework and propose a novel variable-length attention-based mechanism (ACTM) to address it. With ACTM, it is possible to vary the autoregressive order of a TS model over time and model a larger set of probability distributions than with previous approaches. Using this mechanism, we propose a self-supervised DNN architecture for multivariate TS forecasting that learns and takes advantage of the relationships between them. We test our model on two datasets covering 19 years of investment fund activities. Our experimental results show that the proposed approach significantly outperforms typical DNN-based and statistical models at forecasting the 21-day price trajectory. We point out how improving forecasting accuracy and knowing which forecaster to use can improve the excess return of autonomous trading strategies.

Forecasting interval time series using a fully complex-valued RBF neural network with DPSO and PSO algorithms

Direct Interval Forecasting of Wind Speed Using Radial Basis Function Neural Networks in a Multi-Objective Optimization Framework.

A hybrid interval‐valued time series prediction model incorporating intuitionistic fuzzy cognitive map and fuzzy neural network

Direct Interval Forecasting of Wind Power

Learning Interval-Valued Fuzzy Cognitive Maps with PSO Algorithm for Abnormal Stock Return Prediction

Stock price prediction based on deep neural networks

Forecasting Based on High-Order Fuzzy-Fluctuation Trends and Particle Swarm Optimization Machine Learning

Spatiotemporal Adaptive Neural Network for Long-term Forecasting of Financial Time Series

Interval reservoir computing: theory and case studies

A Novel FRBF-Type Model for Nonlinear Time Series Prediction

POD-based Radial Basis Function Neural Network for Nonlinear Time Series Prediction

A BP Neural Network Based on Improved Particle Swarm Optimization and Its Application in Stock Forecasting

A novel interval dual convolutional neural network method for interval-valued stock price prediction

A new hybrid recurrent artificial neural network for time series forecasting

A novel stock forecasting model based on High-order-fuzzy-fluctuation Trends and Back Propagation Neural Network

Multiple-output support vector regression with a firefly algorithm for interval-valued stock price index forecasting

CRPSO-based integrate-and-fire neuron model for time series prediction

A Novel Method Based on FTS with Both GA-FCM and Multifactor BPNN for Stock Forecasting

A Time Series Forecasting Approach Based on Nonlinear Spiking Neural Systems

A New Self-organizing Interval Type-2 Fuzzy Neural Network for Multi-Step Time Series Prediction

Application of neural network based on particle swarm optimization in short-term load forecasting