I noticed some mistranscriptions in the transcript (subtitles, captions) for the course videos. Below is the errata for the 2nd week. I’ll post the errata for the other weeks in those groups.
The format for the errata is as follows. The approximate time is in braces, then the mistranscribed word or phrase follows, followed by a hyphen, then followed by the correct word or phrase. Like this:
[time] mistranscribed word or phrase - correct word or phrase
Video 1 - making recommendations
[2:30] things - thinks
[2:40] raise - rates
[3:02] myself - [?]
Video 2 - using per-item features
[0:01] Recommended - Recommender
[3:32] super strip - superscript
[4:33] On notation - our notation
[5:45] average - algorithm
[8:38] Recommended - Recommender
[10:54] with wish - with which
[11:06] an advance - in advance
Video 3 - collaborative filtering algorithm
[1:32] music - user
[3:20] lists - is
[7:14] cellular algorithm - some other algorithm
[9:24] as y as x - as well as x
[12:46] average - algorithm
[13:39] Recommended - Recommender
Video 4 - Binary labels: favs, likes and clicks
[0:03] Recommended - Recommender
[1:15] nonstop cloud chasers - nonstop car chases
[1:29] average - algorithm
[1:35] daaset - dataset
[2:13] music they - user j
[3:03] a sign - assign
[3:26] assigned to the labor one - assign it the label one
[3:30] assigned to the label zero - assign it the label zero
[3:37] meaningless follows - meaning as follows
[4:46] we’re now - where now
[7:09] some - sum
[8:23] album - algorithm
Video 5 - mean normalization
[0:50] Im ctually - I’m actually
[1:58] b() - b(5)
[2:19] w()•x(i) - w(5)•x(i)
[2:38] hopeful - helpful
[8:11] Recommended - Recommender
[8:44] intensive flow - TensorFlow
Video 6 - TensorFlow implementation of collaborative filtering
[0:18] Hopeful - helpful
[3:34] dissent - descent
[3:44] the rotors - derivatives
[4:43] respective w - with respect to w
[4:59] the computer derivatives - you compute the derivatives
[5:28] collateral - collaborative
[7:37] s tape - as tape
[11:03] TensoFlow - TensorFlow
[11:10] collateral - collaborative
Video 7 - Finding related items
[6:27] steady art - state of the art
Video 9 - Deep learning for content-based filtering
[8:29] console - concert
[9:36] a scale - it scale
[1:19] computational e infeasible - computationally infeasible
Video 10 - Recommending from a large catalogue
[1:19] computational e infeasible - computationally infeasible
[1:21] Recommended - Recommender
[3:56] two cookers - duplicates
[3:57] washed - watched
[4:38] se - say
[5:32] album - algorithm
[6:38] album - algorithm
[6:48] Recommended - Recommender
[7:09] Recommended - Recommender
[7:19] Recommended - Recommender
[7:31] Recommended - Recommender
[7:38] Recommended - Recommender
[7:51] Recommended - Recommender
Video 11 - Ethical use of recommender systems
[1:40] as the most - ads the most
[1:46] as likely - ads likely
[5:48] hope - help
[7:08] recommend the system’s phase - recommender systems face
[8:53] entices - and this is
[9:01] absolute - apps or
[9:08] Apple - apps or
[10:26] the eye - AI
Video 15 - PCA in code
[9:31] You - too
[10:01] hold - help
[10:17] other - older