4 ms·

Nice project, a claver way of skipping repeat write of previous math step work and just do little edit for the NEXT step in the Math homework . github https://

Nice project, a claver way of skipping repeat write of previous math step work and just do little edit for the NEXT step in the Math homework . github https://github.com/jaltekruse/Free-Math https://github.com/jaltekruse/Free-Math On a different note: I think for grading purposes of student math home work which is done on a paper, Deep Learning neural network programs will be advanced enough in 5 years that is by 2023 - we will have a $200 flatbed scanner in every class room, as student finish his school work/home work, she can put the papers on scanner, it will sent scanned image to school server which is running Deep Learning software which Grades and home work and record scores in Database for teacher to view .

jaltekruse 9y ago
I think this might be a way that such a project could be accelerated. Machine learning is good, but math is extremely ambiguous. Symbols are re-used and their orientation relative to other symbols changes their meaning. Some problems want you to simplify, others want you to expand something. This project at least removes symbol ambiguity in written work. Even if you could make a computer understand how to solve a bunch of problems, I'm not certain that blind machine learning would produce a good feedback system for getting students back on course when they misunderstand something. They could probably find some likely candidates where students may have made a mistake, but I don't think anyone has tech anywhere near this right now.
criddell 9y ago
A flatbed scanner? Why not just make an app? The student snaps a photo of the page and errors are overlaid. Every kid has a phone today and I assume that isn't going to change in the next few years. I have kids in high school and 90% of the time when they ask me for help because they are stuck, it's usually due to small arithmetic errors.

jaltekruse 9y ago

I think this might be a way that such a project could be accelerated. Machine learning is good, but math is extremely ambiguous. Symbols are re-used and their orientation relative to other symbols changes their meaning. Some problems want you to simplify, others want you to expand something. This project at least removes symbol ambiguity in written work. Even if you could make a computer understand how to solve a bunch of problems, I'm not certain that blind machine learning would produce a good feedback system for getting students back on course when they misunderstand something. They could probably find some likely candidates where students may have made a mistake, but I don't think anyone has tech anywhere near this right now.

criddell 9y ago

A flatbed scanner? Why not just make an app? The student snaps a photo of the page and errors are overlaid. Every kid has a phone today and I assume that isn't going to change in the next few years. I have kids in high school and 90% of the time when they ask me for help because they are stuck, it's usually due to small arithmetic errors.