Car Price Prediction is most demanded project under machine learning to predict the cost based on many factors like engine type, used/new, year etc. In this data, we will see how other features depend on price by analysing through feature engineering and models.
- Data Understanding
- Feature Engineering
- Feature Selection
- Model Selection
- Prediction of the model
CarAssignment.csv file contains following features:
car_ID,symboling,wheelbase, aspiration, fueltype, CarName, carbody, carlength, doornumber, carwidth, carheight, curbweight, enginesize, boreratio, stroke, compressionratio, horsepower, peakrpm, citympg, highwaympg,price
-
There are 26 features in total, but I have found that few features are not important based on the correlation with the price feature. This enabled for feature selection which are actually influencing the price for prediction.
-
There are categorical features which are not considered for traning because model needs numeric data. So, features like doornumber which are in text format ('two',four') is converted to numeric values.
-
I have considered features of type int and float.Now, the total number features are 16 (after feature engineering).
-
Finally, Linear Regression is applied on final dataset to find the accuracy.
Name: Manasa Noolu
Organization: Becode
email: manasadevinoolu@gmail.com
linkedin profile: (linkedin.com/in/manasanoolu)