E-Commerce Product Recommendation System based on ML Algorithms

TL;DR

This paper presents a machine learning-based recommendation system for eCommerce that personalizes product suggestions using PCA and multiple algorithms, achieving high accuracy and improving user engagement.

Contribution

It introduces a novel eCommerce recommendation model utilizing PCA and four ML algorithms, with RF achieving 99.6% accuracy, enhancing personalization and business outcomes.

Findings

Random Forest achieved 99.6% accuracy

The model reduces features with PCA for better performance

The system improves customer engagement and business benefits

Abstract

Algorithms are used in eCommerce product recommendation systems. These systems just recently began utilizing machine learning algorithms due to the development and growth of the artificial intelligence research community. This project aspires to transform how eCommerce platforms communicate with their users. We have created a model that can customize product recommendations and offers for each unique customer using cutting-edge machine learning techniques, we used PCA to reduce features and four machine learning algorithms like Gaussian Naive Bayes (GNB), Random Forest (RF), Logistic Regression (LR), Decision Tree (DT), the Random Forest algorithms achieve the highest accuracy of 99.6% with a 96.99 r square score, 1.92% MSE score, and 0.087 MAE score. The outcome is advantageous for both the client and the business. In this research, we will examine the model's development and training in detail and show how well it performs using actual data. Learning from machines can change of eCommerce world.