Instance Based Learning

Abstract

This chapter is concerned with the KNN (K Nearest Neighbor) as the simple and practical supervised learning algorithm. As its background, we mention the case based reasoning and lazy learning that underlie in the KNN. Based on the background, we explain in detail the process of carrying out the classifications and regressions of data items by the KNN. We also present the modified versions of KNN as its variants. In this chapter, we assume that the KNN is the supervised learning algorithm, but we cover its unsupervised version in the next part. This chapter is concerned with the KNN (K Nearest Neighbor) as the simple and practical supervised learning algorithm. As its background, we mention the case based reasoning and lazy learning that underlie in the KNN. Based on the background, we explain in detail the process of carrying out the classifications and regressions of data items by the KNN. We also present the modified versions of KNN as its variants. In this chapter, we assume that the KNN is the supervised learning algorithm, but we cover its unsupervised version in the next part.

In Sect. 5.1, we provide the overview of the instance based learning, and in Sect. 5.2, we mention the naive instance based approaches. In Sect. 5.3, we describe entirely the KNN (K Nearest Neighbor) as the most popular instance based learning algorithm. In Sect. 5.4, we mention the KNN variants, and in Sect. 5.5, we make the summarization on this chapter and the further discussions. This chapter is intended to explain the first type of supervised learning algorithms, called instance based learning algorithms.