Skip to content
View elgels's full-sized avatar
2 followers · 30 following

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Block or report elgels

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
elgels/README.md

Hi, I'm Elizabeth Gels

Data Science | Machine Learning | Deep Learning | Time Series Forecasting

Master of Applied Science in Data Science graduate from the Illinois Institute of Technology
Background in analytics, economics, and business strategy.

Featured Projects

Deep Learning — Image Classification

CNN-based image classification project progressing from MNIST (baseline) to CIFAR-10, achieving ~86% test accuracy with detailed error analysis.

  • Convolutional Neural Network (CNN) design and training (PyTorch)
  • Confusion matrix analysis to identify class-level misclassification (e.g., cat vs dog)
  • Performance evaluation across training, validation, and test sets

View Project

Deep Learning — Sentiment Analysis

Transformer-based sentiment analysis project using pretrained BERT embeddings and fine-tuning on the Stanford Sentiment Treebank (SST), achieving improved performance over the earlier CNN-BiLSTM-Attention model with detailed confusion matrix analysis.

  • Fine-tuned pretrained BERT-base Transformer architecture for sentence classification
  • Contextual word representations generated through multi-head self-attention
  • Confusion matrix analysis highlighting challenges in neutral sentiment classification

View Project

Time Series Forecasting — ARIMA/SARIMA

Statistical time series modeling project in R using SARIMA to capture trend and seasonality.

  • Stationarity transformations (log, differencing)
  • ACF/PACF analysis for model identification
  • SARIMA model selection and tuning
  • Residual diagnostics (Ljung–Box, Q-Q plot)
  • Forecasting with prediction intervals


View Project

Machine Learning — Employee Attrition

Analyzes employee data and builds classification models to identify key drivers of attrition and support data-driven HR decisions.




View Project

Technical Skills

Programming

Machine Learning & Statistical Modeling

Supervised & Unsupervised Learning • Statistical Modeling • Forecasting • Time Series Analysis (ARIMA/SARIMA) • Model Evaluation

Deep Learning & NLP

CNNs • LSTMs • Transformers (BERT) • Attention Mechanisms • NLP

Data Visualization

Python (Matplotlib, Seaborn, Plotly) • R (ggplot2) • Tableau

Tools & Technologies

Git • GitHub • Jupyter Notebook • Visual Studio Code • RStudio • BigQuery • MySQL Workbench

Data Handling

Data Cleaning • Data Wrangling • Feature Engineering • APIs • Web Scraping

Additional Projects

CNN-BiLSTM-Attention Sentiment Classification (SST-3) — Deep Learning, Python
Implements a hybrid CNN + Bidirectional LSTM architecture with attention and pretrained GloVe embeddings for sentiment classification and model interpretability analysis.
View Project

IBM Data Science Capstone — SpaceX Launch Prediction — Machine Learning, Python
Predicts Falcon 9 first-stage landing success using classification models and historical launch data.
View Project

Stock Data Extraction — APIs & Web Scraping — Python
Collects and analyzes financial data using APIs and web scraping techniques to explore stock performance trends.
View Project

Bellabeat Case Study — Data Analysis — R
Performs exploratory data analysis on fitness data to generate insights on user behavior and inform business strategy.
View Project

Cyclistic Case Study — Data Analysis — Python, SQL, Tableau
Combines SQL, Python, and Tableau to analyze bike-share data and uncover user patterns to support marketing strategies.
View Project

Professional Certifications

Data Science graduate building projects in forecasting, machine learning, and analytics.

Popular repositories Loading

  1. elgels elgels Public

  2. Salifort-Case-Study Salifort-Case-Study Public

    Data analysis and machine learning project analyzing employee attrition at Salifort Motors. Includes data cleaning, exploratory data analysis (EDA), and predictive modeling (Logistic Regression, De…

    Jupyter Notebook

  3. Bellabeat-Case-Study Bellabeat-Case-Study Public

    Analyzed Fitbit user data using R to identify activity, sleep, and sedentary behavior patterns, and provide data-driven marketing recommendations for Bellabeat.

  4. IBM-Data-Science-Capstone IBM-Data-Science-Capstone Public

    Developed and evaluated machine learning models to predict SpaceX Falcon 9 launch outcomes, using data collection, feature engineering, and classification techniques.

    Jupyter Notebook

  5. time-series-sarima time-series-sarima Public

    Time series forecasting project using SARIMA in R on the AirPassengers dataset, capturing trend and seasonality.

    R

  6. Stock-Data-Extraction-APIs-Web-Scraping Stock-Data-Extraction-APIs-Web-Scraping Public

    Collected and analyzed financial data using APIs and web scraping, with data cleaning and visualization in Python.

    Jupyter Notebook