Beyond size estimation and plan generation, the rest of query optimization is
just comparing costs to choose plans, and relies primarily on stuff you have
done before. 

Size Estimation:

You are given a histogram as follows over an attribute value for which you
additionally record the minimum observed value as 1 and the maximum observed
value as 74.

0-9    4
10-19 40
20-29 70
30-39 60
40-49 22
50-    8

How many occurrences do you expect for attribute value 27?
For value 63?  For value 77? For value 0?  For value 1?  For value 7? 
What is the selectivity of a query that chooses all tuples with value for this
attribute greater than or equal to 35?

Estimate the size of equi-join result for relations R1 and R2, where that join
attribute for each relation has attribute value ranging from 0-100, R1 has
20,000 tuples with 80 distinct values for the join attribute, and R2 has 1000
tuples with 50 distinct values for the join attribute.

Estimate the size of the result if the join above is not an equi-join, but
rather a "theta" join, selecting all tuples where the join attribute value in
R1 is greater than the corresponding value in R2.


Plan Generation:

You have a database with a primary index on ID for Employee and Dept, and
secondary index on Name for the Employee relation only.
You have access methods available for a Nested Loops Join, Nested Indexed
Loop Join, and a Grace Hash Join.  Generate alternative plans for the
following query:

SELECT E.Name
FROM Employee E, Dept D
WHERE E.Dept = D.ID
AND D.Name = "Obfuscation"