Most AI models are developed behind closed doors with unclear documentation and limited visibility into the training data.

One-class support vector machines (OCSVMs) are very effective for semisupervised anomaly detection. However, their performance strongly depends on the settings of their hyperparameters, which has not ...

An improved particle swarm optimization algorithm is proposed, and it is applied to optimize the parameters of support vector machine. The typical Mackey-Glass chaos sequence is predicted through the ...