🎵 RNN-Based Spotify Song Popularity Prediction

This repository contains an implementation of a Recurrent Neural Network (RNN) using PyTorch to predict the popularity of songs on Spotify based on their audio features.

📌 Overview

The goal of this project is to predict the track_popularity score of a song (on a scale from 0 to 100) using various musical and audio features such as danceability, energy, loudness, etc.

This implementation uses a simple RNN model for regression.

🔍 Features Used

The model uses the following numerical features from the dataset:

• danceability
• energy
• loudness
• speechiness
• acousticness
• instrumentalness
• liveness
• valence
• tempo

Target variable:

• track_popularity

📁 Dataset

The dataset file should be named spotify_songs.csv. Make sure to update the path in the script if needed:

df = pd.read_csv(r'c:\faezeh\MachineLearning\ForthPractice\rnn\rnn/spotify_songs.csv')

Missing data is dropped before training.

⚙️ Requirements

• Python 3.8+
• PyTorch
• pandas
• numpy
• scikit-learn
• matplotlib

Install the dependencies with:

pip install torch pandas numpy scikit-learn matplotlib

🏗 Model Architecture

• Input: (Batch, Sequence Length = 1, Features)

• RNN Layer:

  • input_size = 9 (number of features)
  • hidden_size = 64
  • num_layers = 1

• Fully Connected Output Layer: Maps the RNN output to a single regression value (song popularity).

🧠 Training

• Optimizer: Adam
• Loss Function: Mean Squared Error (MSE)
• Epochs: 15
• Batch Size: 64

Training loss is printed per epoch and also plotted at the end.

📈 Evaluation

After training, the model is evaluated on the test set and the final MSE loss is printed.

The training loss over epochs is visualized using a simple line plot.

📊 Output Example

(Use plt.savefig("example_loss_plot.png") in the script if you'd like to save the plot.)

📂 File Structure

├── spotify_songs.csv       # Dataset
├── rnn_regression.py       # Main script (your code)
└── README.md               # Project documentation

▶️ How to Run

python rnn_regression.py