1: \begin{abstract}

2: A nonparametric method to predict non-Markovian time series of partially observed dynamics is developed. The prediction problem we consider is a supervised learning task of finding a regression function that takes a delay-embedded observable to the observable at a future time. When delay-embedding theory is applicable, the proposed regression function is a consistent estimator of the flow map induced by the delay-embedding. Furthermore, the corresponding Mori-Zwanzig equation governing the evolution of the observable simplifies to only a Markovian term, represented by the regression function. We realize this supervised learning task with a class of kernel-based linear estimators, the kernel analog forecast (KAF),

3: which are consistent in the limit of large data. In a scenario with a high-dimensional covariate space, we employ a Markovian kernel smoothing method which is computationally cheaper than the Nystr\"{o}m projection method for realizing KAF. In addition to the guaranteed theoretical convergence, we numerically demonstrate the effectiveness of this approach on higher-dimensional problems where the relevant kernel features are difficult to capture with the Nystr\"{o}m method. Given noisy training data, we propose a nonparametric smoother as a de-noising method. Numerically, we show that the proposed smoother is more accurate than EnKF and 4Dvar in de-noising signals corrupted by independent (but not necessarily identically distributed) noise, even if the smoother is constructed using a data set corrupted by white noise. We show skillful prediction using the KAF constructed from the denoised data. 

4: \end{abstract}

5: