Academic Projects

Academic Resume

This is a collection of Sebastian’s Projects in Data Science and Statistics.

Github: https://github.com/sprestia2009-svg

  • Stock Market Pattern Recognition (In Progress) — Independent Project
    • Abstract: Developing a Time-Series pattern recognition model to identify recurring structural patterns in financial time-series data (Custom Dataset)
      .
  • Symbolic Regression (In Progress) — ASU Research Project
    • Abstract: Developing Symbolic Regression models to discover new mathematical formulas..
  • Mole Detection (In Progress) — Independent Project
    • Abstract: Developing a PyTorch-based CNN architecture to classify benign moles from malignant ones using the ISIC dataset
      .
  • Finding Best Taxi Routes in NYC — Capstone Project
    • Abstract: (my part of team project) Identified the best neighborhoods for NYC taxis through clustering and KDE heatmaps based on a combination of demand, average cost, and trip length (Dataset from Kaggle)
      .
  • Predicting Movie Box Office Revenue with Model — End of Class Project
    • Abstract: Developed a Multiple Linear Regression (MLR) model to predict Movie Box Office using Ridge Regularization; tested model adequacy (Custom Dataset)
      .
  • Impact of the Syrian Civil War on HDI — End of Class Project
    • Abstract: Analyzed changes in Syria’s Human Development Index from pre-civil war and civil war periods; analyzed possible predictors (Dataset from the World Bank)
      .
  • CIFAR-10 & CIFAR-100 Image Classification — Independent Project
    • Abstract: Made PyTorch-based CNN architectures for classification on both the CIFAR-10 and CIFAR-100 datasets
      .
  • MNIST Digit Classification — Independent Project
    • Abstract: Made both a PyTorch-based ANN and a PyTorch-based CNN classifier for the MNIST digit dataset
      .
  • FashionMNIST Classification — Independent Project
    • Abstract: Made both a PyTorch-based ANN and a PyTorch-based CNN classifier for the FashionMNIST clothes dataset
      .
  • Fish Classifier — Independent Project
    • Abstract: Made a PyTorch-based CNN architecture to classify a fish as scary or not (Custom Dataset)
      .
  • ANN from Scratch — Independent Project
    • Abstract: Used the Numpy python module to build an ANN completely from scratch for the MNIST dataset (Custom Dataset)

Note: ANN = Artificial Neural Network; CNN = Convolutional Neural Network (image processing)

.

For more information, check out my LinkedIn.