Low-dimensional approximations of conditional Volterra processes via NPC geometry
By Reza Arabpour, John Armstrong, Luca Galimberti, Anastasis Kratsios, Giulia Livieri
Rating
1590
Battle Count: 91
Relevance
8/10
Highly relevant for modeling and predicting complex financial time series with long memory and stochastic volatility
Implementation Complexity
7/10
Requires understanding of differential geometry and advanced machine learning concepts, but code implementation is provided
Reproducibility
4/5
Code available on GitHub, detailed algorithms and experimental setup provided
About this paper
Methodology: Gaussian Random Projection. Problem types: Dimension Reduction, Time Series Forecasting, Approximation.
The interactive Everscope explorer (charts, battles, favorites) loads below.