SoundSoar

2024 Capstone Results

The study evaluated multiple models, and the highest-performing model type achieved outstanding accuracy. The table below presents the **average performance** of each model type based on multiple runs.

Model Type	# of Models	Avg Accuracy	Avg Precision	Avg F1	Avg Recall
MLPClassifier	1	0.97	0.97	0.97	0.97
HistGradientBoosting	14	0.90	0.90	0.89	0.90
SVM	12	0.90	0.90	0.89	0.90
RandomForest	26	0.90	0.90	0.89	0.90
ExtraTrees	7	0.89	0.89	0.88	0.89
LogisticRegression	12	0.86	0.87	0.86	0.86
LDA	7	0.82	0.77	0.77	0.82
KNN	6	0.80	0.82	0.76	0.80

Table 1 – Summary of Research Results (08/05/2024 - 11/03/2024)

Technical Overview

This section provides a detailed breakdown of the project's technical aspects, including data preprocessing, model selection, and performance evaluation. Each phase of the project, along with its results, is discussed in depth to provide a clear understanding of the methodology and findings.

Demo & Resources

Explore interactive demos, project documentation, and related resources:

Blog – Insights, updates, and discussions on the project
Live Demo (Trending Page) – Explore an example of how you can search for trending songs, as demonstrated in my capstone project.
Project Poster Paper (FLAIRS) – Summary of key findings and insights
Presentation Slides – Overview of the project and its results
Spotify API Documentation – Technical details on data integration
Spotify Privacy Changes – Article discussing recent policy updates (November 27, 2024)