Skip to content

This repository is project submission for Microsoft Engage 2022 program.

License

Notifications You must be signed in to change notification settings

yamini236/Microsoftengage-Used-Cars-Data-Analysis

Repository files navigation

Welcome to Used Cars Analysis Prediction 👋

Predictive Analysis to demonstrate how the Automotive Industry could harness data to take informed decisions.

  1. Why is this analysis required ?
  2. Technologies and Frameworks Used
  3. How is Preproceesing Done?
  4. Analysis of Data
  5. Prediction of Models
  6. Result
  7. Interpretation and Decision
  8. Guide to use Repo
  9. Author

Why is this analysis reuired?

  • User View - To know the price and features of used cars according to your favourable features and helps you decide to buy a new or a used car.
  • Owner View - By observing stats ,gain information about current aspects in automotive industry market and heps to increase production rates and selling units through it.

Technologies and Frameworks Used

  • Built with Jupyter Notebook, Python, and PowerBi.
  • Create predictive models using Machine Learning.

How is Preprocessing Done

  • Missing Vlues - Filled the missing values to generate efficient data.
  • Duplicate Values - Removed Duplicate Values for better prediction and accuracy.

Analysis of data

  • Analysis is done with respect to various features given in train dataset available in M.Engage.ipynb
  • Some of the following are.

Variation of fuels

Fuels


Variation of Transmission with price

Transmission


Variation of Year with price

Year


Famous Cars per Location

Famous


Powerbi report for Additional Analysis as pdf

Powerbi Click here for full pdf pdf


Prediction of Models

  • Prediction of Price of Used Cars is done using Machine Learning Models.
  • Accuracey of different models is compared.
  • Best model is chosen to give accurate result.
  • Initially already provided data from train dataset is used and tested using test dataset.
  • Finally predictions are made using user inputs.

Result

  • Result of the model predictions.

Data from train dataset

f1


f2


f3


Data from User Input

u1


u2


Interpretation and Decision

  • Used Cars are much preferable under low budget especially for college/university students.
  • Diesel driven and Manual mode cars are favoured over others.
  • Volvo series being expensive originally, has a much cheaper cost in case of used cars hence those you love this series can go for it.
  • Selling units and production can be increased efficiently using data analysis.
  • Customers can predict price according to their favourable features.

Guide to Use Repo

Install

git clone https://github.com/yamini236/Microsoftengage-Used-Cars-Data-Analysis.git

Jupyter Notebook

Open M.Engage.ipynb file in jupyter notebook or Google Collab

Select Cells button on top and choose Run all Option.
Run the following Commands
!pip install widgetsnbextension
!jupyter nbextension enable --py widgetsnbextension
!pip install voila
!jupyter server extension enable voila 

Usage

Observe the analysis
To insert new cell choose Insert
To run a new cell Choose Run
To add new package not already installed use - !pip install <packagename>

Author

👤 Yamini Bansal

Show your support

Give a ⭐️ if this project helped you!


This README was generated with ❤️ by readme-md-generator