Experiments with Random Projection
arXiv:1301.3849v1437 citations
Originality Synthesis-oriented
AI Analysis
This work addresses the problem of dimensionality reduction for Gaussian mixture models, but it is incremental as it primarily summarizes and validates existing theoretical findings.
The paper investigates random projection as a dimensionality reduction method for learning mixtures of Gaussians, summarizing theoretical results and demonstrating its effectiveness through experiments on synthetic and real data.
Recent theoretical work has identified random projection as a promising dimensionality reduction technique for learning mixtures of Gausians. Here we summarize these results and illustrate them by a wide variety of experiments on synthetic and real data.