Density estimation of complex data processes by means of neural networks and the integration of these networks in filter methods for the analysis of time series.
This thesis introduces methods to analyse complex data prone to uncertainty and for which posing distributional assumptions is challenging. A fundamental contribution of this thesis is estimating the full distribution of these complex data processes by neural networks. Furthermore, filtering methods are constructed to analyze time series data. Applications are shown for road sensor data.
Peerlings, D. E. W. (2024). Density estimation by neural networks: With applications to non-Gaussian distributions in time series analysis. Dissertation, Maastricht University, doi:10.26481/dis.20241111dp
