en
0.25
0.5
0.75
1.25
1.5
1.75
2
What can machine learning do for open education?
Published on Jun 23, 20142985 Views
One of the big promises of open and massively online education is easy data collection: we can record everything from students’ habits in reading and viewing lectures, to their participation in discus
Related categories
Chapter list
What can machine learning do for open education?00:00
Civilization advances by extending the number of important operations which we can perform without thinking about them01:28
Brain ice berg02:37
Contribution of ML03:16
Why bother?03:58
Goal: Understand how students learn something - 105:23
Goal: Understand how students learn something - 205:56
Example: Geometry tutor06:25
Step-level data07:01
Simplest model: Rasch / I-parameter item response theory - 108:08
Simplest model: Rasch / I-parameter item response theory - 208:47
Structure: Similarity among steps10:30
How principal components analysis got famous11:34
Results of factoring11:48
In our case12:16
Does it work - 113:20
Does it work - 214:41
Structure: Practice makes perfect15:56
What we really want16:44
KC Hypothesis19:08
Compose-by-addition - 120:20
Compose-by-addition - 221:10
Why are some compose-by-addition steps harder? - 121:49
Why are some compose-by-addition steps harder? - 222:27
Hypothesis: Difference is in how much planning is needed23:40
KC discovery24:06
Use data-driven model to redesign tutor24:31
New problems: Isolate practive on planning step24:59
Results25:27
More structure: What's in a KC?26:10
Rule-based cognitive model26:58
Rule-based system27:34
Problem 1: Uncertainty29:39
Rat as bayesian31:17
Quiz - 132:25
Quiz - 232:53
And the rat says....33:34
Bayesian rule learning in classival conditioning36:03
Problem 2: Representation learning36:37
Representation learning38:07
Experiment38:52
New cognitive models are more accurate39:39
Open research question40:12
Summary40:54