This repository contains the trial programs for scrap prediction using the sample dataset. Details about the codes are commented on most of the places to help in better understanding it.
There are two codes:-
- The "Scrap_Prediction_Sklearn.ipynb" is written using the Sklearn and tensorflow frameworks, and various models are tried to get best results.
- The "Scrap_Prediction_python.ipynb" is written in complete python code with some use of numpy, and is a deep learning neural network model whose parameters can be completely changed in the section "Training Of Model" and is expected to five really good results if a minimum of 10000 clean input data examples are fed for training.
Note:- Currently Both the codes are producing results with accuracies of around 70% and are needed to be fed with more data t produce better results