CS 540 Teaching Assistant Information

HW1 : Decision Trees

• Which files should I turn in for the programming section?

You only need to turn in the Java source files (.java). You should delete the .class files from your directory before you submit. If you are using an IDE for development (Eclipse, NetBeans, etc.) make sure your program can be compiled from the command-line. A good way to test this is to copy the files to one of the lab machines. Put all of the .java files in a directory and type the following commands:

javac *.java
java BuildAndTestDecisionTree (parameters)

This should compile and run your program from the command-line. This is how I will be grading your assignment so it's important that this works correctly. If you have problems with this, no problem, just send me an email or come to my office hours and I can help you.


• What should the printed tree format look like?

There is no one set format for this. In general, you want to print out the tree with indentation signifying the node's level in the tree. You should also print out the test that is being performed at each node and the classification if it's a leaf node.

Here's an example of a possible format, where "shape", "age", and "color" are example features and the classifications are '+' or '-':

shape = round:
  age = young:  +
  age = old:
    color = red: - 
    color = green: +
shape = square:
  color = red: +
  color = green: -

It is reasonably easy to write Java code to produce the above using recursion, where a "treeDepth" (an int) is passed in and used to determine the amount of leading spaces per line (say 2 x treeDepth spaces). If you're having trouble with this, come talk to me at my office hours and I can help you out.

HW2 : Searching for Solutions

• What is the heuristic function 'h' measuring?

The heuristic function 'h' represents the estimated cost of getting from a node to the closest goal state. So, if there is a node N in the graph, h(N) is estimating the sum of the arc costs from N to the closest goal state, not the number of hops to the closest goal state.


• For Problem 1, how should I interpret alphabetical order for the goal states?

Assume that the goal states (G1, G2, and G3) come after A, B, C, D, and E, and before S. Within the goal states, assume a numeric ordering, so that G1 comes before G2 which comes before G3.

HW3 : Natural Language Processing

• Which files should I turn in for the programming section?

Follow the same guidelines as for Homework 1 (above). Specifically, I should be able to compile and run your program with the following commands:

javac *.java
java CBR (parameters)
java BayesNet (parameters)

This is how I will be grading your assignment so it's important that this works correctly. If you have problems with this, no problem, just send me an email or come to my office hours and I can help you.


• How should I interpret the ratioXOverY number in the programming section?

You can think of ratioXOverY as the number of observed instances of X divided by the number of observed instances of Y. For instance, if we know that word X appears 16 million times in our corpus (we could determine this number from a search engine query, for example), and Y appears 8 million times, then ratioXOverY = 16 / 8 = 2.0. If you want to calculate P(X) = x / (x + y), you can divide the numerator and denominator by y to get P(X) = (x/y) / (1 + (x/y)) = ratioXOverY / (1 + ratioXOverY). P(Y) would then be one minus this quantity.


• Why are there two copies of each test phrase in the test dataset? Don't I only need the first?

For most of the pairs, the phrases are exactly identical except for the center word. However, this is not always the case. This is because when I prepared the datasets, I ran the POS tagger on both phrases separately in case changing the word changed the predicted part of speech for any of the surrounding words. In certain cases, the parts of speech do change slightly. For CBR, you can't really use this information, since you need to classify a single instance, so you can just use the first phrase for that section. On the other hand, you should use both phrases in the Bayes reasoning part because the change of POS is extra evidence that you can use. If you are using POS in your Bayes reasoning, then for some queries the words in your conditional probabilities for X and Y will be different.

HW4 : Propositional Logic