AOSP-NIT-Surat / Bookstore-ML

The flask server holding Image Processing algorithm.
MIT License
0 stars 2 forks source link

[Analysis] Rate books condition based on images #2

Open GrayFlash opened 3 years ago

GrayFlash commented 3 years ago

This issue is for you to add various possible approaches in docs, with the proper link to citations.

Suggest more features around it.

shivambmishra10 commented 3 years ago

sir i want to confirm it that in this issue we have to pic the image of front page of book and then it our ml model would predict the rating , that only the issue is or something else?

GrayFlash commented 3 years ago

Yup, and for this issue you basically explain the approach you might want to take, and code is not required for this one.

shivambmishra10 commented 3 years ago

1.Sir according to my knowing here we can use a fundamental algorithm logistic regression because we want output in discreate form such as categorical form and via logistic we get output in this form only as per sigmoid function work in fram of probability and then setting cutoff we will convert it into discreate one. 2.Next for image manipulation basics of open cv would work but not very sure. 3.then we will train the model on the basis of the data we have available and return output. This is what i think. So please review it.

vivektangudu2024 commented 3 years ago

I Think we have to train our model on 5 different datasets as we want to rate from 1-5, we will prepare a data set which contains ratings like which books come under 1 ,2.. now our model will relate this to output the grading

GrayFlash commented 3 years ago

yup can do that, but is there any data set for the same??

vivektangudu2024 commented 3 years ago

i think to find a data set is difficult so we will give user to fix rate but max rate is MRP .customer who want to buy can have a chat with him and decide their price .