The advantages and disadvantages of the machine learning combat-k-nearest neighbors algorithm

The K approximation algorithm is an instance-based learning, and when using the algorithm we have to have an example training sample data close to the classification results.

Advantages: High accuracy, insensitive to outliers

Disadvantages:

• Time complexity and space complexity are relatively large. (If the training sample dataset is large, it takes a lot of space to hold the data, and it takes time to predict the data and train the sample data set for each piece of data.) ）
• The infrastructure information for any data cannot be given, so it is not possible to know the characteristics of the average instance sample and the typical case sample.
• The classifier could not be persisted.

The advantages and disadvantages of the machine learning combat-k-nearest neighbors algorithm