A comparison of multitask and single task learning with artificial neural networks for yield curve forecasting
作者：
Highlights：
• Novel use of neural networks and multitask learning for yield curve forecasting.
• Multilayer perceptron using the most relevant features achieved the best results.
• The most relevant features depend on target yield and forecasting horizon.
• Synthetic data from linear regression model tends to improve forecasting accuracy.
• Encouraging results for the development of robust forecasting systems for bond market.
摘要
•Novel use of neural networks and multitask learning for yield curve forecasting.•Multilayer perceptron using the most relevant features achieved the best results.•The most relevant features depend on target yield and forecasting horizon.•Synthetic data from linear regression model tends to improve forecasting accuracy.•Encouraging results for the development of robust forecasting systems for bond market.
论文关键词：Machine learning,Neural network,Multitask learning,Yield curve forecasting,Yield forecasting,Bond market
论文评审过程：Received 20 March 2018, Revised 26 September 2018, Accepted 6 November 2018, Available online 6 November 2018, Version of Record 12 November 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.11.012
