2.3 - Give specifics for percepts, actions, goals (and performance measure), and environment (note 2.5 for an example)
Chapter 3
For each of the six basic uninformed search strategies, describe a realistic search problem where that strategy would be more appropriate than the other 5, and state why.
Chapter 4
4.1 and 4.2 - Give at least one reasonable heuristic for each problem
Chapter 6
Review 6.2, do 6.3 (prove some directly using logic rules instead of a truth table) and 6.10
Chapter 7
7.2, 7.6 (just first paragraph), 7.10, and 7.14 (give a 1 page discussion)
Chapter 9
9.1, 9.4, 9.5, 9.8
Chapter 11
11.1, 11.4, 11.7a-c
Chapter 14
14.2, 3 (give an exact probability using the normalization technique)
Chapter 15
15.1
Chapter 18
18.3
Given the training set below and using the ID3 algorithm, show carefully (work out the proper numbers) which attribute would be the first attribute chosen to partition the training set. Use only information gain as the criteria.
Example/Year/GPA/Marital Status/Output
1 Fr Hi M N
2 So Lo S P
3 Se Med M P
4 So Hi M N
5 So Hi M N
6 Fr Med S P
7 Se Hi M N
8 Fr Lo S N
9 Se Lo S P
Chapter 19
19.1 (use threshold logic units)
Given a 3-input threshold unit (1 if > 0, else 0), plus a bias input, initial weights of 0, and the following training set, show the learning sequence followed until convergence (or a maximum of 3 epochs). Assume a learning rate of .1. Show your results in a tabular format with the following entries.
Pattern /Target /StartingWeight Vector/ Net/ Output/ Vector of weight changes
a)
1 0 0 -> 0
0 0 0 -> 0
0 1 1 -> 1
1 1 1 -> 1
b)
.2 -.3 .6 -> 1
.3 0 -.2 -> 0
-.5 -.1.1 -> 1
Chapter 22
22.7