Welcome to my Capstone Project, The Song Stats App Featuring Predictive Modelling! This project combines data analysis, machine learning, and integration with Spotify's API to create an interactive app using Streamlit. The objective is to analyze The Song Popularity Dataset, ustilize Machine Learning Modelling to predict song popularity, and to combine those into a Steamlit app that allows users to interact with the data.
-
Predictive Modeling Presentation π½οΈ:
- In the
presentation
folder, you'll find a PDF showcasing my predictive modeling presentation. This includes insights, analysis steps, and an overview of the Streamlit app. - The presentation details the three machine learning models used: k-Nearest Neighbors, Logistic Regression, and Random Forest.
- In the
-
Data π’:
- The
data
folder contains both the raw and cleaned datasets obtained from Kaggle.com, named 'The Song Popularity Dataset'.
- The
-
Code π:
- The
code
folder hosts my Jupyter notebook, providing insights into the data cleaning, exploratory data analysis, and model fitting/testing processes. - Explore the steps I took in analyzing The Song Popularity Dataset.
- The
-
Images π·:
- The
images
folder stores visuals generated by my Streamlit app. More specifically, these photos are mainly album music covers generated from the Spotify API, which are then stored into this folder.
- The
-
Streamlit App π±:
- The
app.py
file is the main script using Visual Studio Code, running the Streamlit app. Interact with the app to explore musical attributes, visualize and explore data, get song recommendations, and predict song popularity.
- The
-
Additional Scripts π₯οΈ:
polarplot.py
: Contains code for generating polar plots.songrecommendations.py
: Utilizes Spotify's API to provide song recommendations based on user input.
-
Predictive Modeling Presentation: π£οΈ
- Explore the
presentation
folder to view the PDF of the predictive modeling presentation.
- Explore the
-
Data Analysis and Modeling π:
- Delve into the
code
folder to find my Jupyter notebook detailing data cleaning, exploratory data analysis, and machine learning model implementations.
- Delve into the
-
Raw and Cleaned Data π’:
- Access the
data
folder to obtain both the raw and cleaned versions of 'The Song Popularity Dataset'.
- Access the
-
Streamlit App π²:
- Run the Streamlit app by executing
app.py
in Visual Studio Code.
- Run the Streamlit app by executing
-
Additional Scripts π:
- Check out
polarplot.py
for polar plot code andsongrecommendations.py
for Spotify API-based recommendations.
- Check out
Ensure you have the necessary requirements by referencing the requirements.txt
file.
Feel free to explore, contribute, and enjoy The Song Popularity App! π΅