dmubook-errata
Errata to Decision Making Under Uncertainty: Theory and Application
These items have been fixed in the second printing and later.
- Alg. 4.12 (line 4): replace ; with | to be consistent with the notation used in this section (Jon Cox)
- Eq. 6.11: switch a and s in the transition function to be consistent with the notation in the book (Max Egorov)
- Eq. 2.61: (v_i - \hat\mu) should be squared
- P. 59, sentence before eq. 3.3: replace u^1 with c^1 (David Wagner)
- Sec. 3.1.1: in independence condition, second \succ should be \succeq (Ellen Blaine)
- Fig. 3.6: The code used to generate this figure had a small bug. The histograms should be more distributed as shown here.
- Sec. 4.7.1 (second to last sentence): drop "and m" (Hamza El-Saawy)
- Fig. 6.6: some of the alpha vectors did not print correctly (noted by Yi-Chun Chen); the figure should look like this:
- Sec. 6.4.5 (second sentence): replace "each belief state B" with "each belief state in B" (Zongzhang Zhang)
- Ch. 7: references 12, 15, 38, and 63 should have "POMDPS" and "MDPS" changed to "POMDPs" and "MDPs" (Zongzhang Zhang)
- Eq. 6.31: should be
$\sum_s b(s) \sum_{s'} O(o \mid s', a)T(s' \mid s, a)$ (Junzi Zhang)
These items will be fixed in the 6th printing:
- Alg. 4.1: U should have \pi as a superscript everywhere (Benjamin Antin)
- Eq. 4.24 and 4.27: The rightmost Ta should be transposed (Tim Wheeler)
- Alg 2.1, line 2: Drop the subscript on the s
- Alg 4.9, line 13: for consistency arg max should be over A(s) rather than all of A
- Eq. 4.36: It should be explained that if N(s,a) = 0, then the exploration bonus is treated as infinity (Tim Wheeler)
- Eq. 6.15-6.17: The Kalman filter equations should look like this (Tim Wheeler)
- Table 10.2: ft/s in the rightmost column should be ft/min (Cem Polat)
- Sec. 1.5:
[10]should be swapped with[11](Vladislav Ankudinov) - Sec. 4.7.3: two instances of
$P(\lambda; \theta)$ should be$P(\lambda \mid \theta)$ for consistency (Ross Alexander) - After eq. 4.2: "is worth" should be "are worth" (Brad Rafferty)
- Immediately before eq. 5.31: the
$n$ should be$m$ to be consistent with eq. 5.31 (Kyu-Young Kim) - First sentence of 5.2.2: change "Q" to "U"
- Sec. 4.3.1 last sentence: "four" should be "five" (Preston Rogers)
- Eq. 4.25: The T_s immediately after the inverse should be T_a (Allan Li)
- Alg. 6.2: sampling state from b should be within the repeat loop so lines 4 and 5 should be swapped (Tim Wheeler)
- Eq. 6.1: drop the t subscripts (Kyu-Young Kim)
These items will be fixed in a future printing:
- For consistency, algorithms in Ch. 7 should use U instead of V.