Amazon SageMaker Now Supports k-Nearest-Neighbor and Object Detection Algorithms
Amazon SageMaker now supports the k-Nearest-Neighbor (kNN) and Object Detection algorithms to address additional identification, classification, and regression use cases in machine learning. This addition expands the list of built-in algorithms for SageMaker to 15.
The kNN algorithm can be used to address classification and regression problems. As an example, the classification of an unlabeled image can be determined by the labels assigned to its nearest neighbors. This is useful for recommendation systems, anomaly detection, and image/text classification. For regression problems, kNN can be used to predict a number based on the function of the labels of its neighbors, which is typically set as the average or the median.
Object detection is the process of identifying and classifying objects in an image. With the new Object Detection algorithm in Amazon SageMaker, you can more easily build and train models capable of detecting multiple objects in an image during inference. Bounding boxes are placed around identified objects and then the objects are classified.
Support for kNN and Object Detection algorithms in Amazon SageMaker is now available in the US East (N. Virginia), US East (Ohio), US West (Oregon), Europe (Ireland), Europe (Frankfurt), Asia Pacific (Tokyo), Asia Pacific (Seoul), and Asia Pacific (Sydney) AWS regions.