Olist Brazilian E-commerce Analysis

Introduction

Explore insights from 100k orders (2016-2018) across Brazilian marketplaces. From data loading to model deployment, this project covers EDA, preprocessing, modeling, NLP for customer satisfaction, and customer segmentation.

Data Source

Olist Kaggle Dataset

Deployed App

Streamlit App

Test Data

Two sample CSV files (EDA.csv, Clustering Sample.csv) are provided in the repository for EDA and clustering analysis testing.

Table of Contents

1. Introduction
2. Data Loading
3. Data Cleaning
   • Merging Dataframes
   • Handling Missing Values
   • Drop Duplicates
   • Feature Engineering
4. Exploratory Data Analysis (EDA)
   • Univariate Analysis
   • Multivariate Analysis
5. Data Preprocessing
   • Data Encoding
   • Feature Scaling
   • Handle Imbalance
6. Modeling
   • Apply ML Models
   • Hyperparameter Tuning
7. Pipeline
8. NLP for Customer Satisfaction
9. Customer Segmentation
   • RFM Analysis
   • K-Means
10. Model Deployment (Classification & Clustering)
11. Wrap Up & Conclusion