KMeans Clustering Part 1 - Determining The Optimal Number Of Clusters

Описание: In this video I'm going to walk you through how to determine the optimal number of clusters in a data set for a KMeans cluster analysis in R with various libraries in RStudio. KMeans is one of the most reliable and used methods for cluster analysis currently in use in data science and data analysis. This clustering is used often times to determine and gain insights on customers, products sold, services, etc.

Get a free stock like Apple, Exxon Mobil, GM, Ford, Tesla or Facebook worth up to $500 just by signing up and joining Robinhood using my link: https://join.robinhood.com/davidm16707

To do an accurate and reliable KMeans cluster analysis we need to first start with determining the optimal number of clusters in our data. So this is part one of a three-part series where I will walk you through the entire realm of clustering with KMeans. In the end you will be able to document the number of clusters, why you chose that number, four methods to back up your number, you will be able to do the cluster analysis and then append the cluster data back to your data set and then you'll be able to create wonderful, insightful maps with this cluster data.

So basically I am going to walk you through a complete and proper KMeans analysis in R (if you aren't familiar with R or RStudio please watch my other videos on loading data and starting with R and RStudio on my YouTube channel). In this video we will strictly deal with loading the correct libraries, loading your data set, cleaning up your data, placing it in a matrix and scaling it and then determining the optimal number of clusters from your data. We will use the elbow within sum of squares method along with several other methods.

I hope you found this interesting and helpful. Clustering is used every day in data science projects to determine proper customer segmentation, product segmentation and service segmentation. It is also used in interesting projects like determining disease outbreaks, flu transmission rates, financial forecasting, business cycles, financial analysis and more.

Thank you for watching please take a moment to subscribe and like. Also be sure and check out the other two parts for this series and all the other great data science data analysis videos I have on my YouTube channel.

Thanks again and God bless!