Q1. What are the four essential requirements of a distance measure as might be used in cluster analysis, multidimensional scaling (MI)S) or the Kohonen SOM?
Q2. In a histogram, explain what are represented by the area and the height of the histogram bars.
Q3. What is the problem of overplotting in data visualization? Describe briefly two way in which we can overcome this difficulty.
Q4. Explain how a star plot can be used to visualize a set of variables.
Q5. Explain the difference between multiplicative and additive time series.
Q6. What is the difference between Metric and Non-metric Multidimensional Scaling?
Q7. Make a sketch to show how a matrix scatter plot is constructed.
Q8. Explain how external validation can be carried out for cluster analysis.
Q9. What is the difference between exploratory data analysis (EDA) and confirmatory data analysis(CDA)?
Q10. Using a sketch of a scatter plot and the example of Pearson's Correlation Coefficient, r, explain what is meant by Simpson's Paradox.