feat(exponential): add moments strategy

[xImoZA]

This PR introduces the Method of Moments estimation strategy, providing a generalized and analytical implementation for the Exponential distribution.

Changes

• Generalized Strategy: Implemented the base moments_strategy, which allows for extensible implementations and raises a NotImplementedError for any distribution that does not have a registered strategy.

Exponential Implementation:

Added the analytical solution for Exponential distribution parameters (loc and rate) based on weighted empirical moments. The implementation handles various parameter configurations:

• Fixed loc -> Updates rate based on the first moment.
• Fixed rate -> Updates loc based on the first moment.
• Non-fixed parameters -> Updates both using variance and mean.

Testing: Added comprehensive unit tests covering:

• Base dispatcher behavior using a DummyDistribution.
• Parameter recovery for all fixing combinations.
• Numerical stability checks and overflow handling.