mdnazrulislam-code/carpriceprediction

CarPricePrediction

About Data

This data is collected from the Kaggle . You will get this dataset from here "https://www.kaggle.com/datasets/balaka18/quikr-cars-scraped"

Data Quality

• year column has many no-year values
• year object is object
• price column has ask for price string values
• price columns is object
• kms_driven object to int
• kms_driven has nan values
• fuel_type has nan values
• keep first 3 words from name column

Performed Steps

• Data Cleaning,
• Data Processing,
• EDA
• Features Extraction,
• One Hot Encoding using column_transformer from scikit-learn
• Making Pipeline for the categorical features
• Model Selection
• Run 1000 times of for loop to get the best r2_score
• build website using flask
• deployed the model with streamlit website

Want to check real time ?

Clone the repository

https://github.com/mdnazrulislam-code/CarPricePrediction.git

STEP 01- Create a conda environment after opening the repository

conda create -n py310 python==3.10 -y

conda activate py310

STEP 02- install the requirements

pip install -r requirements.txt

# Finally run the following command
streamlit run main.py
python application.py

Now,

open up you local host and port