Unsupervised Learning is a machine learning method that tries to find the hidden structure/pattern among unlabeled data. The important thing here is that unsupervised learning is used when we do not have any information about the label of the data we have. In this project, I developed an unsupervised learning model using the USArrest dataset.
USArrest is a data set that gives the number of crimes committed in the states of America, according to their categories. This dataset is very suitable for implementing a machine learning project with an unsupervised learning model because it consists of unlabeled data.
In this project, I interpreted the data and found a relationship between them and used Kmeans clustering method. I visualized the clusters I associated and analyzed them in the scatter plot.
But how many clusters should it consist of? At this point, I found the optimum number of clusters with the Elbow Method. In the following stages, I applied the hierarchical clustering and PCA steps and made the fit process.