PyCaret Project to B and Deploy an ML App using Streamlit Business Objective Customers are the lifeblood of any firm. Customer retention should be implemented with the firm's long-term goals in any organization that serves a large number of customers. In general, businesses spend a lot of money to attract new consumers, but often don't know that their existing clients account for the majority of their sales and earnings. Customer analytics, which serves as the foundation for all marketing efforts, includes + approaches such as predictive modeling, data visualization, information management, and segmentation. Customer segmentation and a product recommendation engine are employed effectively to maximize sales and conversions. We shall concentrate on customer segmentation in this project. Marketers can use segmentation to better customize their marketing efforts to different audience segments. Customer segmentation is the process of breaking customers into multiple groups based on shared criteria in order for businesses to promote each group efficiently and appropriately. We'll be working on an end-to-end project in which we'll create a model with the help of the Pycaret library, Streamilit, Github, and the Streamlit cloud. PyCaret is a Python-based low-code machine learning library that allows you to go from data preparation to model deployment in minutes in your preferred notebook environment. In this particular project, K means clustering will be used to segment the data. We shall then build a Streamlit application and deploy it on the Streamiit cloud by linking it to the Github repository containing our project, Data description The dataset that we will use for this project is a part of PyCaret datasets and is called ‘jewellery’. It comprises customer profiles with columns such as age, income, spending score, and savings, and we're aiming to classify customers based on these so that our marketing team can target potential product buyers. To build an unsupervised learning model with the help of PyCaret library in python that can categorize the customers into different segments and deploy the model using streamlit.Tech Stack Language - Python Libraries - pycaret, pandas, streamiit Approach 1. Importing the required libraries and packages 2. Open the config file. (This is a configuration file that can be edited according to your dataset) Get the dataset Setup PyCaret environment Model Creation Model Assigning Plotting model Making predictions Saving Model 10. Creating streamlit application 11. Creating a github repository for the project 12. Connecting streamiit cloud to github 13. Deploying the project OONBARY Modular code overviewinput Lconfig.yami Liewel_data.csv sre Lengine.py LmLpipetine Lutils.py Lprocessing.py Lmodel.py Lstreamtit_app Lapp.py Lrequirements.txt fib Customers Segmentation with Pycaretipynb output Final_kmeans_model.pki Ltestesy Once you unzip the modular_code.zip file you can find the following folders within it 1. input 2. ste 3. output 4, lib 1) input folder - It contains all the data that we will need for analysis. * A config file, with some basic configuration parameters which can be edited according to your dataset. © A jewel_data.csv file, that has 505 rows of customer data with 4 columns as- Age, Income, Spending score and Savings. 2) src folder - This is the most important folder of the project. This folder contains all the modularized code for all the above steps in a modularized manner. This folder consists of:4) * engine.py © mlpipeline The ml_pipeline is a folder that contains all the functions put into different python files which are appropriately named. These python functions are then called inside the engine.py file © streamlit_app This folder contains the python file for the streamlit application that will be used for deployment of this project. © requirements.txt This file will help you install all the packages that are required to run the project successfully. You can install theses packages by using the ‘command — pip install -r requirements.txt, output folder - The output folder contains model that we trained for this data. This model can be easily loaded and used for future use and the user need not have to train all the models from the beginning, lib folder - This is a reference folder. It contains the original ipython notebook that ‘we saw in the videos. Project Takeaways 1 eNonsen 9. 10. 11. 12. 13. Understanding customer segmentation and its usage Understanding the working of PyCaret How to perform data preprocessing using PyCaret? How to build a model using PyCaret? Understanding streamiit and its usage Understanding the importance of deploying a model Learning how to productionize a model How to build a streamlit app? Understanding what is github Understanding github repositories and branches How to create a pull request in github? How to create a feature branch in github? Lear to create a github repository for your project14. Understanding the steps to connect streamilit cloud to github 15. Understanding the project deployment process using streamlit