Ai Study Buddy banner

Ai Study Buddy

4 devlogs
11h 42m 38s

Updated Project: a Flask web app that recommends study times, topics, or review strategies based on user input and past learning behavior. its not actually ai or a studdy buddy it just sounded catchy (but you can choose little pets)
Nothing has c…

Updated Project: a Flask web app that recommends study times, topics, or review strategies based on user input and past learning behavior. its not actually ai or a studdy buddy it just sounded catchy (but you can choose little pets)
Nothing has changed since my last ship beside a few minor bug fixes i just forgot to make a devlog before i ship last time so I didnt log half my hours.

This project uses AI

Github copilot for debugging and asking questions

Demo Repository

Loading README...

sanadeghat

Shipped this project!

Hours: 6.51
Cookies: 🍪 60
Multiplier: 9.2 cookies/hr

Same as last time beside some minor fixes i just forgot ot make a devlog before my last ship so I didnt log like 6 hours.

sanadeghat

FIANLLY FINISHED IT
the app has structure, CSS and login now yipeeeeee and ive spoent the past ,WAY TOO LONG like 6 hours hyepr focusing on thuis bc i rly didnt wanna do my hoemwork but im like so proud of her

Attachment
0
sanadeghat

Shipped this project!

Hours: 5.15
Cookies: 🍪 50
Multiplier: 9.78 cookies/hr

Ok so last tiem i made a super basic flask app it was just a very basic machine elkarning mdoel and a dform this time i wanted to take taht and do MOREEEE. you can log in (this was teh feature taht i dreaded the most but it turned out fine), log study sessions, tell it your assigments nd see a graph of your study lately. ALSO THE PART THAT I MMOS TPROUD OF IS YOU CAN CHOOSE LITTLE PETS ISNT THAT THE CUTEST(thank you awsome font). the thing that i would change for next time is insread of json files i uwould use data bases but that was maybe too many new thinsg for this project.

sanadeghat

Today i tried diffrent moderls beside decision tree classifier and tested their (SEEMED VERY SUSPICIOUS but apprently i just didnt have enough data hetes how accurate we looking now and pretty bc css). I also had to find an afficient away to generate enough data to train my modal on.

Attachment
0
sanadeghat

Changed how i recive data, now you can choose between 3 energy levels
added some css
added a place were you can try out hypothatical study sessions to see if they would be “Effective”
gives reccomendations

Attachment
0
sanadeghat

Currently the app can:
take in study sessions data pooints from 4 set subjects (the ability to adfd sujescts will be added later)
train a modal based on tah data
recoomend ideal tiem t ostudy (the accuracy is a bit TBC but yk it does soemthing)
it has a predictionb form where you give it a “planned study session” and it tells you if its effective or not

most notbly i learneda bunch of new scikit learn functions like .predict and .feature_importances_

Attachment
Attachment
2

Comments

Sam Wu
Sam Wu 2 months ago

Looking forward to seeing this later!

scitechexperiments

cool!!