Used cars price predictor

Mar 2021 - Jun 2021

Associated with Shaheed Zulfikar Ali Bhutto Institute of Science and Technology

Problem Statement

Building a model for Pakistan used car market with the best accuracy by understanding the factors on which the pricing of used cars will depend. Variables are significant in prediction as it influences the actual worth of the car and how well those variables affect the price but from the perspective of a seller, it is difficult to predict the price of a used car accurately. So, based on the data our aim is to use machine learning algorithms to develop good models for predicting used car prices.

Data Cleaning Techniques

To make the dataset ready for the machine learning algorithm, that data needs to be cleaned before fitting it to the model. The steps that we took to clean our data are as follows:


LabelEncoder and OneHotEncoder Techniques were used for Data Cleaning and optimization. Unencoded columns were dropped so there was no null, undefined, or object value present in the test and training dataset.

If need then the data will be derived from the actual data Now split the data into training and testing sets.

After performing the above data cleaning techniques our data is ready to get fitted in the machine learning algorithm.