Understand the fundamentals of Principal Component Analysis (PCA) and identify opportunities to combine variables.
Conduct correlation testing with various sets of variables in Google Sheets.
Combine highly correlated variables, visualize the data, and consider next steps in Google Sheets.
Survey data sets are often deceptively complex because surveys collect a wide variety of data covering a wide variety of topics and experiences. To further the complexity of survey data, the respondents answering the questions come from a wide variety of backgrounds and stages in their customer journey. It is reasonable that it would be a challenge to boil down survey data into actionable insights because it can be deceptively complex. With large sets of data, Principal Component Analysis or PCA is a useful tool that reduces and transforms variables to a leaner form that allows for a speedier analysis. In this project you will gain hands-on experience with the principles of Principal Component Analysis using survey data. To do this you will work in the free-to-use spreadsheet software Google Sheets. By the end of this project, you will be able to confidently apply Principal Component Analysis concepts to transform large sets of variables into a leaner set of data that still contains the most relevant information. Note: This course works best for learners who are based in the North America region. We’re currently working on providing the same experience in other regions.
Familiarity with spreadsheet software, factor analysis, and correlation testing. "Design a Factor Analysis Using Survey Data" is recommended.
Review the fundamentals of Principal Component Analysis (PCA) and combining variables.
Identify use cases for PCA and refine variable selection for the project.
Access Google Sheets, import survey data, and examine variables that are likely correlated.
Identify variables of interest and conduct a correlation test.
Compare results and review the process of correlation testing.
Combine highly correlated variables, create a visualization, and consider next steps.
Access the ClustVis webtool for visualizing clustering and multivariate data.
Build a PCA model with Heart data and run a Principal Component Analysis
Compare results and review PCA with multivariate data from multiple sources and interpret the findings in ClustVis.
I feel that this project was a little slow, it took I think 10 minutes to actually begin working. This is best suited for beginners, as it covers very basic materials.
Deeper knowledge on PCA should be introduced, more complex project or case should be given. Learning PCA through Excel is never enough.
Course was very helpful. However it would be better if analysis on more number of variables could be conducted
我能直接通過 Web 瀏覽器來完成此指導項目，而不必安裝特殊軟件嗎？