Updated codebase

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,52 @@
+<h1 align="center">Medical Insurance Cost Prediction</h1>
+
+<div align= "center">
+  <h4>Medical Insurance Cost Prediction using Random Forest Regressor.</h4><br>
+  <img src="https://github.com/SahilChachra/Medical-Cost-Prediction/blob/master/sampleImages/doc.gif">
+</div>
+
+&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
+
+## :innocent: Motivation
+
+To predict things have been never so easy. I used to wonder how Insurance amount is charged normally. So, in the mean time I came across this dataset and thought of working on it! Using this I wanted to know how few features determine our insurance amount! 
+
+## :hourglass: Some Screenshots
+
+![](https://github.com/SahilChachra/Medical-Cost-Prediction/blob/master/sampleImages/Cor.png)
+
+![](https://github.com/SahilChachra/Medical-Cost-Prediction/blob/master/sampleImages/AgevsCharges.png)
+
+![](https://github.com/SahilChachra/Medical-Cost-Prediction/blob/master/sampleImages/deployments.png)
+
+## :star: Features
+<ol>
+    <li>Exploring the dataset</li>
+    <li>Converting Categorical values to Numerical</li>
+    <li>Plotting Heatmap to see dependency of Dependent valeu on Independent features</li>
+    <li>Data Visualization (Plots of feature vs feature)</li>
+    <li>Plotting Skew and Kurtosis</li>
+    <li>Data Preparation</li>
+    <li>Prediction using Linear Regression</li>
+    <li>Prediction using SVR</li>
+    <li>Prediction using Ridge Regressor</li>
+    <li>Prediction using Random Forest Regressor</li>
+    <li>Performing Hyper tuning for above mentioned models</li>
+    <li>Plotting Graph for all Models to compare performance</li>
+    <li>Preparing model for deployment</li>
+    <li>Deployed model using Flask</li>
+</ol>
+
+## :key: Results
+
+####  Model gave 86% accuracy for Medical Insurance Amount Prediction using <code>Random Forest Regressor</code>
+
+
+## :file_folder: Dataset
+The dataset used can be downloaded here (Kaggle) - [Click to Download](https://www.kaggle.com/mirichoi0218/insurance)
+
+## :heart: Owner
+Made with :heart:&nbsp;  by [Sahil Chachra](https://github.com/SahilChachra)
+
+## :eyes: License
+MIT © [Sahil Chachra](https://github.com/SahilChachra/Medical-Cost-Prediction/blob/master/LICENSE)

README.md

@@ -41,6 +41,17 @@ To predict things have been never so easy. I used to wonder how Insurance amount
 
 ####  Model gave 86% accuracy for Medical Insurance Amount Prediction using <code>Random Forest Regressor</code>
 
+## How to run?
+<ol>
+    <li>Clone the repository</li>
+    <li>Check Python3 version (by typing python3 in terminal) and if required update it to 3.12</li>
+    <li>Install dependenceis - pip3 install -r requirements.txt</li>
+    <li>Download the insurance.csv file if not present in the folder by default. (Kaggle link below)</li>
+    <li>Now, open Medical Cost Prediction notebook and work on your model</li>
+    <li>In the end the model will be saved with the name - "rf_tuned.pkl"</li>
+    <li>Now, after trainig you model, simple run - python3 app.py to load Web app</li>
+
+</ol>
 
 ## :file_folder: Dataset
 The dataset used can be downloaded here (Kaggle) - [Click to Download](https://www.kaggle.com/mirichoi0218/insurance)

requirements.txt

@@ -1,37 +1,7 @@
-backcall==0.1.0
-certifi==2019.11.28
-Click==7.0
-colorama==0.4.3
-cycler==0.10.0
-decorator==4.4.1
-Flask==1.1.1
-Flask-Cors==3.0.9
-graphviz==0.13.2
-ipython>=7.16.3
-ipython-genutils==0.2.0
-itsdangerous==1.1.0
-jedi==0.15.2
-Jinja2==2.11.3
-joblib==0.14.1
-kiwisolver==1.1.0
-MarkupSafe==1.1.1
-matplotlib==3.1.2
-numpy==1.22.0
-pandas==0.25.3
-parso==0.5.2
-pickleshare==0.7.5
-prompt-toolkit==3.0.2
-pydotplus==2.0.2
-Pygments==2.7.4
-pyparsing==2.4.6
-python-dateutil==2.8.1
-pytz==2019.3
-scikit-learn==0.22
-scipy==1.4.1
-six==1.13.0
-sklearn==0.0
-traitlets==4.3.3
-wcwidth==0.1.8
-Werkzeug==0.16.0
-wincertstore==0.2
-gunicorn==20.0.4
+flask
+numpy
+pandas
+matplotlib
+seaborn
+scikit-learn
+xgboost