In the Dynamic Classifier Selection (DCS), we provide a list of machine learning models. Each model is trained with the training data. When a new prediction needs to be made, we select the model that can make the best prediction.

We can take the following two approaches in the Dynamic Classifier Selection:

• We can use Overall Local Accuracy (OLA)
• We can use Local Class Accuracy (LCA)

In our previous article, we discussed Dynamic Classifier Selection (DCS) with Overall Local Accuracy (OLA). In this article, we are going to discuss Dynamic Classifier Selection (DCS) with Local Class Accuracy (LCA).

In DCS with LCA, we provide a list of models. We train each model with the training data. When new data comes, each model makes the prediction. After that, we select k nearest neighbors of the new data from the training data. And we evaluate each model based on that. In other words, each model makes a class prediction for the k nearest neighbors. And based on those predictions, the performance of each model is evaluated. After that, the model that makes the best prediction is selected.

Dynamic Classifier Selection (DCS) with Local Class Accuracy (LCA) using Python

We can use the following Python code to implement Dynamic Classifier Selection (DCS) with Local Class Accuracy (LCA). …