Showing Revision 1 created 09/11/2014 by Cogi-Admin.

1. タイトル：
2. 概説：

3. 290 0:00 - 0:01
   HI, I'm a Mac.
4. 1350 0:01 - 0:04
   >> [LAUGH] No. >> All right, we'll start over.
5. 3780 0:04 - 0:05
   Let's do this right.
6. 4600 0:05 - 0:06
   Hi, I'm Charles Isbeil.
7. 6200 0:06 - 0:07
   >> True.
8. 7100 0:07 - 0:08
   >> I'm a professor at Georgia Tech.
9. 8450 0:08 - 0:09
   >> True.
10. 9130 0:09 - 0:11
    >> This is Michael Littman.
11. 10520 0:11 - 0:13
    >> True. >> He is also a professor at Georgia Tech.
12. 13060 0:13 - 0:14
    >> False.
13. 14040 0:14 - 0:17
    >> No, that's actually true, while you're a professor at Brown,
14. 16510 0:17 - 0:18
    you're also an adjunct professor here at Georgia Tech.
15. 20150 0:20 - 0:22
    I'm a professional donut thief.
16. 21590 0:22 - 0:22
    >> False.
17. 22190 0:22 - 0:24
    >> That is correct I am not a professional.
18. 24020 0:24 - 0:26
    Together, we are teaching the course Machine Learning.
19. 26110 0:26 - 0:27
    >> True.
20. 27000 0:27 - 0:30
    >> The lectures will only be available on prime number calendar days.
21. 30070 0:30 - 0:31
    >> False.
22. 31190 0:31 - 0:33
    >> The mini course is on supervised learning.
23. 33140 0:33 - 0:34
    >> True.
24. 34010 0:34 - 0:38
    >> Supervised learning itself was invented in the 1830s by Gauss.
25. 37750 0:38 - 0:39
    >> False?
26. 38710 0:39 - 0:39
    >> Actually, I don't really know.
27. 39450 0:39 - 0:42
    It sounds like the sort of thing you'd invent but, probably not,
28. 41680 0:42 - 0:43
    so let's go with that.
29. 43250 0:43 - 0:46
    Okay, supervised learning is the problem of learning to map inputs to
30. 46320 0:46 - 0:48
    predictions of true or false.
31. 48500 0:48 - 0:49
    >> False.
32. 49420 0:49 - 0:52
    >> Good one, it also includes other kinds of predictions such as regression,
33. 52040 0:52 - 0:54
    where the output might be vectors or numbers.
34. 54190 0:54 - 0:55
    >> True.
35. 54990 0:55 - 0:58
    >> It's an important component of all sorts of technologies from stopping credit
36. 57950 0:58 - 0:59
    card fraud.
37. 58540 0:59 - 1:01
    >> True. >> To finding faces in camera images.
38. 60660 1:01 - 1:03
    >> True. >> To making taste your breath mints.
39. 63070 1:03 - 1:04
    >> False?
40. 64050 1:04 - 1:07
    >> I have no idea, but that sounded ridiculous so let's say false.
41. 67232 1:07 - 1:08
    To recognizing spoken language.
42. 68350 1:08 - 1:09
    >> True.
43. 69220 1:09 - 1:12
    >> Our goal here is the give you the skills that you need to recognize how to
44. 72090 1:12 - 1:13
    apply these technologies.
45. 73280 1:13 - 1:14
    >> True.
46. 73910 1:14 - 1:15
    >> And for interpreting their output, so
47. 75340 1:15 - 1:17
    that you can solve a range of data science problems.
48. 77300 1:17 - 1:18
    >> True.
49. 78070 1:18 - 1:20
    >> And for surviving a robot uprising.
50. 79670 1:20 - 1:21
    >> False. >> No, that is absolutely true.
51. 80830 1:21 - 1:25
    The most important thing you could get out of this course is learning how to
52. 84659 1:25 - 1:27
    survive the upcoming robot uprising.
53. 86712 1:27 - 1:29
    Okay, very good.
54. 88880 1:29 - 1:31
    You got an accuracy of 85%.
55. 91062 1:31 - 1:32
    >> True.
56. 91767 1:32 - 1:35
    >> Wrong, though after you answer this question,
57. 94873 1:35 - 1:37
    you will have an accuracy of 85%.
58. 99699 1:40 - 1:41
    >> True.
59. 100590 1:41 - 1:43
    >> That is correct, but false also would have been correct.
60. 103070 1:43 - 1:44
    Congratulations.