Real-time inflation forecasting using non-linear dimension reduction techniques

In this paper, we assess whether using non-linear dimension reduction techniques pays off for forecasting inflation in real-time. Several recent methods from the machine learning literature are adopted to map a large dimensional dataset into lower-dimensional set of latent factors. We model relationship between and factors constant time-varying parameter (TVP) regressions with shrinkage priors. Our models then used forecast monthly US The results suggest that sophisticated yield forecasts highly competitive linear approaches based on principal components. Among considered, Autoencoder squared components have high predictive power one-month- one-quarter-ahead inflation. Zooming performance over time reveals controlling relations data is particular importance during recessionary episodes business cycle or current COVID-19 pandemic.