update README

README.md

@@ -113,7 +113,7 @@ The following describes how to use this repository in your own implementation.
 A submodular function can be represented as an object of a class implementing one of two interfaces, depending on whether the maximization algorithm can use marginal increments to speed things up:
 
 1. General submodular functions should implement the [`AbstractSubmodularFunction` interface](./submodmax/abstract_optimizer.py).
-2. Submodular functions for which the maximization algorithm only computes marginal increments, for which we can use a computational trick to speed things up. I.e., if the submodular objective function is sequentially evaluated, we might only need to take into account the difference between the sets between two consecutive evaluations, and update the previous function value using this information. In this package, we refer to this as **function value reuse**. These submodular functions should implement the [AbstractSubmodularFunctionValueReuse` interface](submodmax/value_reuse/abstract_optimizer.py).
+2. Submodular functions for which the maximization algorithm only computes marginal increments, for which we can use a computational trick to speed things up. I.e., if the submodular objective function is sequentially evaluated, we might only need to take into account the difference between the sets between two consecutive evaluations, and update the previous function value using this information. In this package, we refer to this as **function value reuse**. These submodular functions should implement the [`AbstractSubmodularFunctionValueReuse` interface](submodmax/value_reuse/abstract_optimizer.py).
 
 #### General submodular functions - evaluation without function value reuse
 Without function value reuse, submodular functions can be implemented similar to the [basic example](https://github.com/joschout/SubmodularMaximization#basic-example), which can also be found in [./examples/example_Andreas_Krause.py](./examples/example_Andreas_Krause.py)