soachishti / SENNE

Code and datasets for paper "Nearest Neighbor Ensembles---An Effective Method for Difficult Problems in Streaming Classification with Emerging New Classes"

Code of Nearest Neighbor Ensembles---An Effective Method for Difficult Problems in Streaming Classification with Emerging New Classes

Files

Scripts:

• demo: script that you can directly run on synthetic dataset
• getData: script that turn normal dataset into stream data with known classes and new class, and split dataset into training set and test set
• model_SENNE: function of build models
• Predict: function of predicting classes of test dataset
• modelupdate: function of updating models using data in the buffer
• calc_score: function of calculating New Class Score
• hyperSphere: function of building hypersphere clusters
• plot_error: plot results intuitively

Data:

The .mat files contain all datasets used in the paper.