Loading [a11y]/accessibility-menu.js
Exploring the Impact of Feature Data Normalization and Standardization on Regression Models for Smartphone Price Prediction | IEEE Conference Publication | IEEE Xplore

Exploring the Impact of Feature Data Normalization and Standardization on Regression Models for Smartphone Price Prediction


Abstract:

Smartphone has become a necessity and the most accessible for every individual and is mainly used for communication. Smartphone price is higher than the previous years be...Show More

Abstract:

Smartphone has become a necessity and the most accessible for every individual and is mainly used for communication. Smartphone price is higher than the previous years because of the new features provided. Smartphone Price Prediction will be using machine learning to predict the smartphone price with the provided feature. The dataset will be obtained through web-scraping with 14 features that consist of categorical and numerical features. Categorical features consist of Brand, Chipset, CPU, GPU, Screen Type, Resolution, and Total Camera, while numerical features consist of RAM, Weight, Storage, Screen Size, Pixel Density, battery capacity, and Price for the target. The model that is proposed for the study will be a decision tree regressor, random forest regressor, k-nearest neighbor regression, and multilayer perceptron regression using two pre-processing data methods that is data normalization and data standardization. The result of the research is data normalization is the most impactful pre-processing data on regression models in terms of performance and error rates. Random forest regressor with no scaling performs better than other models with a score of 0.10425 and 0.97552 for MAPE and R2 respectively.
Date of Conference: 24-25 August 2023
Date Added to IEEE Xplore: 17 October 2023
ISBN Information:

ISSN Information:

Conference Location: Malang, Indonesia

References

References is not available for this document.