Instanzenbasierte

Instanzenbasierte is a German term that translates to "instance-based" in English. It refers to a paradigm or approach where reasoning, decision-making, or problem-solving relies directly on specific examples or instances from past experience. Instead of abstracting general rules or models, an instance-based system learns from and operates on the stored data points themselves.

In machine learning, instance-based learning, also known as memory-based learning, is a category of algorithms that defers the generalization step until a new query is made. When a new data point needs to be classified or predicted, the algorithm compares it to the stored training instances. The outcome is determined based on the similarity between the new instance and its nearest neighbors in the training set. A well-known example of an instance-based learning algorithm is the k-Nearest Neighbors (k-NN) algorithm.

The core idea behind instance-based approaches is that similar situations are likely to have similar outcomes. By maintaining a collection of past experiences (instances), the system can leverage this memory to address new, similar situations. This can be particularly effective when the underlying concept is complex or difficult to represent with a concise model, or when the data is noisy and abstracting a rule might lead to inaccuracies. The trade-off is often increased computational cost during the query phase, as extensive comparison with stored instances is required.