INSURANCE PREMIUM PREDICTION USING MLOPS

Problem Statement:

The goal of this project is to give people an estimate of how much they need based on their individual health situation. After that, customers can work with any health insurance carrier and its plans and perks while keeping the projected cost from our study in mind. This can assist a person in concentrating on the health side of an insurance policy rather han the ineffective part.

Approach:

The classical machine learning tasks like Data Exploration, Data Cleaning, Feature Engineering, Model Building and Model Testing. Try out different machine learning algorithms that’s best fit for the above case. Some Famous Algorithms: - Multiple Linear Regression, Decision tree Regression and Gradient Boosting, Decision tree, Regression

Results:

We have to build a solution that should able to predict the premium of the personal for health insurance

DEPLOYED APPLICATION LINK:https://mlops-application.herokuapp.com/

FOR DOCUMENTATION OF THIS PROJECT VISIT HERE:DOCS

STEPS ARE MENTIONED BELOW FOR MAKING THE ENTIRE PIPELINE

step1 :

conda create -n filename python=3.7

step2 : Create template.py file and create files and directories

step 3: Inside the params.yaml file understand it and append the code i have written
Also same for dvc.yaml
These 2 are the brain and heart for this mlops dvc project

step 4:
Inside the src dir create get_data.py
OR
touch src/get_data.py
The main objective for making this file is to get track of params.yaml file and to read the data present in our local system.If you want to read data from s3 bucket or azure you can also customize the code and fetch the data

step 5:
After step 5 open up vscode or pycharm terminal
Follow these commands
1.pip install -r requirements.txt
2.git init
3.dvc init
4.dvc add data_given/csv_file_name.csv
5.git add .
6.git commit -m "committed"
7.git remote add origin git repo https address
8.git branch -M main
9.git push origin main

step 6:
Inside the src dir create load_data.py file
OR
touch src/load_data.py
The main objective for this file is we will write a function which will fetch .csv file and append to raw data folder.

step 7:
Inside the src dir create split_data.py
OR
touch src/split_data.py
The main objective for this file by the sklearn library we will divide the data into 75:25 ratio and then append to processed folder

step 8:
Inside the src dir create train_evaluate.py
OR
touch src/train_evaluate.py
This is the favorite step for all data science enthusiasts we will train and evaluate the model here.

step 9:
After evaluating the model i got around 85% accuracy of r2 score and around 0.07 normalized rmse score formula for r2=1-rss/tss and rmse formuala np.sqrt((x-x^)^2)/n and normalized rmse formula rmse/max-min of the target.Save and dump it to saved_models directory.

step 10:
Come to terminal and follow these commands dvc repro dvc params diff if not worked try below one dvc metrics show This will track all the scores,parameters of ml algorithm and in future if u changed any ml algorithm it will track and show in logs.

step 11:
Come to terminal and install few packages pip install pytest pip install tox These packages are used for testing environment.

step12:
create file named tox.ini it is used bcz tox is a generic virtualenv management and test command tool that is used for. -->>Helps in running your tests in each environment. schema

[tox]
    envlist=py37
    [testenv]
    deps=pytest
    command=pytest -v

step 13:
make dir test inside test create files __init__.py,conftest.py,test_config.py These files help us for testing environment.

step 14:
come to terminal pytest -v type it will test and give test cases pass or fail

step 15: create setup.py in root dir it is help us to create python package. Come to terminal Type tox it will install all the dependencies if no error it will pass the test cases otherwise it will give error.

step 16: name="src" and create package of our model Come to terminal python setup.py sdist bdist_wheel it will create .tar file and create package src

step 17: All our procedure fininshed Now time to create webapp

├───static
│   └───css
│           main.css
│
└───templates
        404.html
        base.html
        index.html

step 18: app.py on root dir for creating flask api Now make routes like \ for rendering home page and /predict for rendering predictions and extra predict_postman you can add here for testing in postman which is jsonify it will give json results

step 19: create package for this project so that new user can install our package and can use this project
visit here : https://packaging.python.org/en/latest/tutorials/packaging-projects/

step 20: For automation of the project create dir .github\workflow\ci-cd.yaml we used here github actions for automating our project

Cloning this project

git clone https://github.com/rohanpatankar926/internship_project.git
cd project directory pip install -r requirements.txt python app.py
All set now you cloned this project successfully
Author:Rohan patankar
For any queries related to ml/dl contact me rohanpatankar926@gmail.com

Thank You