3. K-Nearest Neighbor algorithm

K Nearest neighbor (k-nearest NEIGHBOR,KNN) Classification algorithm

1, definition: If a sample in the feature space in the k nearest (that is, the closest feature space) of the sample most belong to a category, then the sample belongs to this category.

2, calculation formula:;

3, K-Nearest neighbor algorithm needs to do standardized processing;

4. K-Nearest Neighbor algorithm API

5. Advantages:

1) simple, no parameter handling, no training required

6. Disadvantages:

1) lazy algorithm, when the test sample classification of large computational capacity, memory overhead;

2) must be specified k value, K value is not properly selected classification accuracy is not guaranteed;

7, the use of the scene: small data volume, thousands of ~ Tens of thousands of samples.

8, speed up the search speed-based on improved algorithm kdtree.

3. K-Nearest Neighbor algorithm