1: \begin{abstract}

2: A solution manifold is the collection of points in a $d$-dimensional space satisfying a system of $s$ equations with $s<d$. 

3: Solution manifolds occur in several statistical problems including

4: missing data, 

5: algorithmic fairness,

6: hypothesis testing, 

7: partial identifications, and nonparametric set estimation. 

8: We  theoretically and algorithmically analyze solution manifolds. 

9: In terms of theory, we derive five useful results:

10: smoothness theorem,  stability theorem (which implies the consistency of a plug-in estimator),

11: convergence of a gradient flow,

12: local center manifold theorem

13: and convergence of the gradient descent algorithm.

14: We propose a Monte Carlo gradient descent algorithm to numerically approximate a solution manifold.

15: In the case of the likelihood inference, we design a manifold constraint maximization procedure to

16: find the maximum likelihood estimator on the manifold.

17: \end{abstract}

18: