1: \begin{abstract}

2: This paper considers the problem of reconstructing missing parts of functions based on their observed segments. 

3: It provides, for Gaussian processes and arbitrary bijective transformations thereof, theoretical expressions for the $L^2$-optimal reconstruction of the missing parts. 

4: These functions are obtained as solutions of explicit integral equations. 

5: In the discrete case, approximations of the solutions provide consistent expressions of all missing values of the processes. 

6: Rates of convergence of these approximations, under extra assumptions on the transformation function, are provided. 

7: In the case of Gaussian processes with a parametric covariance structure, the estimation can be conducted separately for each function, and yields nonlinear solutions in presence of memory.

8: Simulated examples show that the proposed reconstruction indeed fares better than the conventional interpolation methods in various situations. 

9: \end{abstract}

10: