Identify Weather Pattern using k-means algorithm.
Following activity is part of the Machine Learning With Big Data by University of California, San Diego through Coursera.
This activity guides through the process of performing cluster analysis on a dataset using k-means using Apache Spark. The cluster analysis was performed on the minute-weather.csv dataset using the k-means algorithm. The minute-weather.csv dataset contains weather measurements such as temperature, relative humidity, etc., from a weather station in San Diego, California, collected at one-minute intervals. The goal of cluster analysis on this data is to identify different weather patterns for this weather station.
From elbow-plot, it was clear that the elbow in the curve is between 10 and 15, so for this analysis k = 12 was choosen.