Review3: (09/25 - 10/09)

• What are the differences between unsupervised and supervised discretizer.
• Which one should we run first? Why?
• What’t the advantages of using supervised discretizer?
• What should we do when the supervised discretizer have multiple dependent variables?
• Is mean or median a better metric to compare two learners? Why?
• What can you tell me about these two data samples?

 one         * --|            , 0.28,  0.29,  0.32,  0.41,  0.48
 two             |    -- * -- , 0.71,  0.79,  0.80,  0.90,  0.98

How would you rank the following five results from five treatments. Show all working.

 x1  0.34  0.49  0.51  0.6
 x2  0.6   0.7   0.8   0.9
 x3  0.15  0.25  0.4   0.35 
 x4  0.6   0.7   0.8   0.9 
 x5  0.1   0.2   0.3   0.4

Consider the following table.

• How are effect size tests different to significance tests?
• How are parametric tests different to non-parametric tests?
• Explain: parametric tests are faster than non-parametric tests since they make limiting assumptions.
• Define the cohen test.
• Explain, with a diagram, ttests how much normal curves overlap. In your explanationd draw (a)two normal curves which tests would say are not significantly different and (b) two more normal curves that ttests would say are signfificantly different.
• Write down the pseudocode of cliffsDelta.
• In bootstrapping, what is a bootstrap sample? How does bootstrapping using 100s of such samples to determine if two lists are significantly different.

When examining data from multiple sources, chant this mantra:

• Visualize the data, somehow.
• Check if the central tendency of one distribution is better than the other; e.g. compare their median values.
• Check the different between the central tendencies is not some small effect.
• Check if the distributions are significantly different;

For each of the above four points,

• List two ways to achieve all those goals (hint, some of your answers come from the above).