Gaussian models and evaluation of clusters
For the sample model of distinguishing numbers, different clustering models are used and the most suitable dividing point (number of clusters) and the suitable covariance matrix structure are found. Choosed a suitable Gaussian model and used four methods to explore whether the clustering can be successfully divided or which ones are classified into one category. According to sample analysis, it may be difficult to successfully use clustering to make detailed divisions, or the accuracy may be relatively low. Explored the impact of matrix selection or model changes on sample analysis.
This was done as a part of my masters group assessment