Multi-view data(Scene15,ALOI,HandWritten,BDGP,MNIST-USPS,Prokaryotic,yale_mtv,100leaves)

This dataset collection aggregates eight widely used multi-view benchmarks: Scene15, ALOI, HandWritten, BDGP, MNIST-USPS, Prokaryotic, yale_mtv, and 100leaves. Together, they cover diverse domains including natural scenes, generic objects, handwritten digits, biological and genomic data, human faces, and fine-grained plant species. Each sample is described by multiple heterogeneous feature views with different dimensionalities and statistical properties, and the datasets vary in the number of classes, sample sizes, and view configurations. This collection provides a comprehensive testbed for studying multi-view representation learning, clustering, and related tasks under realistic settings with diverse semantic categories and feature types.