AI class errata: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
Line 248: | Line 248: | ||
| memorisation of a criterion | | memorisation of a criterion | ||
| minimization of a criterion | | minimization of a criterion | ||
|- | |||
| [http://www.youtube.com/watch?v=_-Ol1cXIWvQ _-Ol1cXIWvQ] | |||
| 1:40 | |||
| tests would show | |||
| tests should show | |||
|} | |} |
Revision as of 10:59, 4 November 2011
These are my notes on errata I've discovered in the AI class coursework.
Videos
Video | Time | Is | Should be |
---|---|---|---|
4G5mH4FW-WY | 0:51 | curvatic | quadratic |
0RmqLOxexh4 | 0:26 | corresponding | closed-form |
0RmqLOxexh4 | 0:41 | interation | iteration |
0RmqLOxexh4 | 0:54 | up with it | update it |
rAcwpZJqAZA | 0:46 | local minimum | local minima |
dKKigX6nhyU | 0:05 | quadratic arrow | quadratic error |
R1o9wbhnv94 | 0:43 | plausible | closed-form |
yOSGC67bOIk | 0:50 | sets | data sets |
yOSGC67bOIk | 1:53 | grade descent | gradient descent |
yOSGC67bOIk | 1:57 | grade descent | gradient descent |
yOSGC67bOIk | 3:10 | and the error is zero, then no update occurs | then the error is zero, and no update occurs |
xRf9wAeU1kI | 1:43 | robust-ness | robustness |
xRf9wAeU1kI | 2:00 | integer | iterative |
xRf9wAeU1kI | 3:32 | so that just | plotted just |
xRf9wAeU1kI | 3:49 | Map | Mapped |
xRf9wAeU1kI | 4:36 | to write | to derive |
xRf9wAeU1kI | 4:44 | These messages | These methods |
ZLEilYyt28c | 0:35 | condition probabilities | conditional probabilities |
PoRpuj4bijU | 0:00 | is an easy answer | is easily answered |
tOSoqfK9UNE | 1:05 | If your graph input | If you graph your input |
tOSoqfK9UNE | 1:35 | they are to be one | there ought to be one |
tOSoqfK9UNE | 1:36 | your're | you're |
kFwsW2VtWWA | 0:04 | are seen not to be completely random determinants | seem not to be completely randomly determinate |
EZEOXNFgu8M | 0:08 | interpretively assume | typically assume |
EZEOXNFgu8M | 0:38 | structure and data | structure in data |
EZEOXNFgu8M | 1:21 | drawing signal | joint signal |
W2dkDmHFMWg | 2:04 | They were derived | They will be derived |
zaKjh2N8jN4 | 0:18 | found interatively | found iteratively |
zaKjh2N8jN4 | 0:34 | Euclidian | Euclidean |
zaKjh2N8jN4 | 1:28 | has attained the center | is attained at the center |
myqnyxkdQpc | 0:17 | corresponding step | correspondence step |
myqnyxkdQpc | 1:10 | local minimum | local minima |
3zlXl82LUVI | 0:02 | one interation | one iteration |
_DhelJs0BFc | 0:44 | horizontal access | horizontal axis |
_DhelJs0BFc | 2:04 | then it is | than it is |
_DhelJs0BFc | 2:08 | periphery summary over here | periphery somewhere over here |
rMcw3uu4efY | 1:55 | complete the derivative for spectrum mu | compute the derivative with respect to mu |
rMcw3uu4efY | 2:03 | we can still get this | we instead get this |
rMcw3uu4efY | 2:11 | next to zero | it's still zero |
rMcw3uu4efY | 2:56 | stresses internal | is just its internal |
pRGEQy7BgiY | 0:04 | multivariant | multivariate |
mlz-1yfyeoU | 0:06 | the fit from data | how to fit them from data |
1CWDWmF0i2s | 0:47 | Their movement is smooth away | They move in a smoother way |
tTr7547zVCc | 0:07 | sum of all possible | sum over all possible |
tTr7547zVCc | 0:45 | should we call | which we will call |
TFViJ3P6NwM | 0:12 | specifically M1 sigma | specifically mu and sigma |
DODedtJZ3FA | 0:17 | first situation | first iteration |
_-Ol1cXIWvQ | 0:40 | memorisation of a criterion | minimization of a criterion |
_-Ol1cXIWvQ | 1:40 | tests would show | tests should show |