Comparison Of Model-Based Clustering to Other Clustering Methods: an Example on Meteorological Data

Using a methodology comprised of model-based clustering and panel data analysis, we have tried to draw conclusions from real-life meteorological based on philosophy science. In this study, used consisting yearly observations annual mean temperatures gathered the 58 stations in regions Turkey compared results other agglomerative methods that were derived upon an earlier study aforementioned data. We then configured clusters as separate dummy variables effects which was put test by longitudinal analysis. The more successful according between R^2 obtained analysis but best comparative demonstrated climate change occurs differently across Turkey.