The World of Computing

CS100

Haverford College

Spring 2000

Examination 3

Sample Key

  1. Briefly define, in the context of this course, each of the following terms.
    • nonfeasible - a problem which can be specified and solved, but the resources (i.e., time) needed to solve the problem increases much faster than the size of the problem (e.g., exponentially), meaning we can only solve small problems with little impact, more meaningful problems could not be solved for any practical purposes (e.g., TSP)
    • unsolvable (a.k.a., undecidable) - a problem which can be specified but not solved (e.g., Halting Problem).
    • TSP - Traveling Salesperson Problem, given a set of cities to visit, find a tour which visits each city only one time, ends at the start city, and completes the tour for minimal cost.
    • expert system - a field in AI which concentrates on a narrrowly defined problem with a limited set of rules and tokens which works like an expert in that problem.
    • analog - measuring, as oppossed to discrete, which counts.
    • multiplexor/decoder - a multiway switch to select one of N lines; since the machine is binary, lg N select lines are needed to select among N lines.
  2. Explain why searching the phone book for a given phone number is a more difficult problem than searching for a given name. For extra credit, provide the time complexity (i.e., the function of N which describes how much time it takes to solve the problem) for each of these searching problems, where N is defined as the number of names - phone numbers in the phone book.
    • The names are ordered so we can use binary search (O(lg N)), while the numbers are not sorted, requiring sequential search (O(N)).
  3. Discuss why a deterministic "intelligent" machine would find it difficult to resolve the following statement as either true of false: "There are no absolute rules -- one should never say never."
    • A contradiction arises when this statement is applied to itself, since it explicitly does what it professes should not be done.
  4. Provide at least three features generally agreed to determine if an entity is "intelligent".
    • ability to learn, reconcile inconsistent statements, guess when needed, perform as an expert in certain areas.
  5. Briefly discuss the distinction(s) between reasoning and behavior pertaining to the definition of intelligence. Into which camp does the "Turing Test" fall?
    • Reasoning looks at the very processes that result in intelligent behavior, while behavior only looks at the external result of an intelligent agent.
    • The Turing Test would fall in the latter category, defining intelligence as looking for behaviors attributed to intelligent humans.
  6. Add in binary:
    0 1 0 1 + 0 1 1 1 = 1 1 0 0
  7. Name the two actions that a register can perform with data, and give a brief description of each action.
    • reading - copying the value without changing it
    • writing - changing the value
  8. What is the name of the component responsible for interconnecting the processor and memory (as well as other devices)?
    • The bus.
  9. Consider the circuit from the sample exam ...
    • provide the corresponding truth table
      • a b c
      • 0 0 0
      • 0 1 1
      • 1 0 1
      • 1 1 0
    • b. provide an equivalent symbolic expression (i.e., boolean function)
      • c = (ab)' (a + b)
    • c. what common function is this circuit equivalent to? XOR

Page maintained by John Dougherty, Computer Science, Haverford College