Dennis Dang

Hi there! I’m Dennis.

This website is a collection of personal projects, case studies, and maybe a few notebooks of me tinkering around with data on my free time.

Projects

Automated Chorus Detection
Deep Learning Audio Analysis Digital Signal Processing
Automated Chorus Detection
A Convolutional Recurrent Neural Network (CRNN) pipeline to extract audio features, segment features into musical meters, and predict whether the meters belong to the chorus.
View the project repository
Audio Segmentation Using Chroma Features and Self-Similarity Matrices
Digital Signal Processing Audio Analysis
Audio Segmentation Using Chroma Features and Self-Similarity Matrices
This project explores the application of Chromagram Self-Similarity Matrices (SSMs) for segmenting music into its structural components such as intro, verse, chorus, and outro.
View the notebook
Case studies
Springboard Data Science Case Studies
Case studies
I can segment more than just music! Here are some case studies that cover a wide range of industry-relevant problems, such as customer segmentation and purchasing predictions, using techniques like clustering, decision trees, and regression to extract business-critical insights.
View case studies
Mixin - The Quest for an A.I. DJ
Mixin Python Deep Learning MIR
Mixin - The Quest for an A.I. DJ
A DJ and mixtape generator that uses deep learning, Python, and a sprinkle of music theory. (This is an ongoing project)
Learn more about this project
Mixin Exploratory Data Analysis (EDA)
Mixin Data Visualization Feature Engineering Audio Analysis
Mixin Exploratory Data Analysis (EDA)
This EDA explores the Mixin dataset, comparing audio features from Spotify with my analysis, testing tempo estimation and beat/meter quantization methods, and visualizing audio features with chorus labels to inform feature selection for modeling.
View the notebook on Github
Novelty-based Section Boundary Detection
Mixin Audio Analysis Audio Segmentation
Novelty-based Section Boundary Detection
Segments audio using RMS energy calculations, tempo estimation, beat tracking, CQT, MFCC extraction, novelty curve computation, and peak detection to extract features, visualize music structure, and detect section boundaries in audio signals.
View the notebook on Github

About Me

I’m a data scientist with a passion for leveraging data and machine learning to drive positive change and create cool things. My journey is shaped by my diverse background in research, education, technology, and psychology, with a sprinkle of music on the side.

I hold a BA in Psychology from the University of Pennsylvania, where I was heavily involved in undergraduate research through independent studies, conference participation, and research assistantships at various labs like Dr. Angela Duckworth’s Character Lab.

After completing my undergraduate studies, I joined Dr. Carol Connor PhD and the Individualizing Student Instruction Lab (ISI Lab), where I further developed my programming and data science knowledge through my work on various classroom learning technologies. I later joined the UC Irvine School of Education as a doctoral student specializing in Educational Data Science. My research focused on leveraging machine learning on data from EdTech products to improve student outcomes and educational equity.

While I ultimately decided to pause my PhD studies due to personal circumstances related to the COVID-19 pandemic, my time at UCI further fueled my passion for data science and its potential to drive meaningful change and discoveries. Currently, I am a Data Science Fellow at Springboard, where I’m leveraging my academic background and knowledge to tackle industry-relevant problems as I transition from the world of research to the realm of applied data science.

Music has always been a significant part of my life, and I’ve found ways to combine my love for music with my data science interests. I enjoy DJing and creating mixes on Soundcloud, and I even create audio/visual novels from my mixes. I use my passion for music as motivation to develop quantitative skills by treating music as data (i.e. vectors/matrices/signals that can be manipulated and analyzed). So not only am I learning, but I’m also creating cools things like Mixin, an AI-powered DJ and mixtape generator.

Outside of my data science and music projects, I enjoy dancing, longboarding, and playing pickleball.

Education

Data Science Fellow - Springboard
12/2023 - Present

Currently enrolled in intensive 6-month program mastering Python, SQL, statistical analysis, and machine learning, with 1:1 mentorship from an industry expert.

  • Applied quantitative techniques and experimental design to business-relevant problems in various case studies.
PhD Candidate in Educational Data Science - UC Irvine
2020-2021
  • Completed the first year of the program before deciding to pause my studies due to personal/family health issues related to the COVID-19 pandemic.
  • My research focus was on the application of machine learning techniques on EdTech data (e.g. Learning Management Systems) to improve student outcomes and educational equity.
  • I was a Graduate Student Researcher on the UCI-MUST grant where we focused on analyzing students' online behavior and engagement within the Canvas Learning Management System, the primary platform used for UCI courses, in order to gain a deeper understanding of how students interact with course materials, participate in discussions, and complete assignments within the online learning environment.
BA in Psychology - University of Pennsylvania
2018
View my undergraduate thesis on Humility
  • GPA: 3.6/4.0
  • I was awarded the South Asian Studies Summer Research Grant to conduct ethnographic research in Nepal in 2017, which laid the foundation for my senior thesis.
  • I developed a novel psychometric scale for Humility in my senior thesis, which received the prestigious John P. Sabini Research Award from Penn’s psychology department.
  • I served as an undergraduate research assistant for Dr. Angela Duckworth’s Character Lab and Dr. Ayelet Ruscio’s Boundaries of Depression and Anxiety Lab.
AA in Social and Behavioral Sciences - Irvine Valley College
2016
  • GPA: 4.0/4.0
  • I was highly active in Psi Beta, the national honor society in psychology, and participated in numerous undergraduate research projects. I designed and presented an original psychology experiment titled, “What’s Better than Learning?: Hedonic Contrast in Education” at the following research conferences:
    • Bay Honors Symposium, University of California, Berkeley - May 2016
    • Annual Convention of the Western Psychological Association, Long Beach, CA - April 2016
    • Honors Transfer Council of California, University of California, Irvine - March 2016
    • Southern California Conferences for Undergraduate Research, Harvey Mudd College - November 2015
    • Irvine Valley College/ Saddleback College Research Symposium, Irvine Valley College - November 2015 (recipient of the Outstanding Abstract Award)