Welcome To My Portfolio
Table of Contents
- Machine Learning/Deep Learning Projects
- Data Analysis
- Other Python Projects
- Front End/Web Development
- Tutorials
Machine Learning/Deep Learning Projects
Flower Classification - Triple Model Classification (Keras)
Flowers are classified in this project, and the performance of three different models is compared. A custom model, as well as ResNet50 and InceptionV3 are compared.
Text Classification Comparision - (Keras)
Text classification is carried out on multiple bodies of text. An embedding model is used for the classification, and the performance of an LSTM model is compared as well.
X-Ray Classification Transfer Learning Tutorial - (Keras)
Classification of X-Ray medical images is carried out in this project. One classifier is trained on the original dataset and then a second set of data has perturbations applied to it to simulate image corruption. An example of transfer learning is carried out after this, retraining the original model on the new dataset.
CIAFR-10 Classification - Transfer Learning - (PyTorch)
Classification done on the CIFAR-10 dataset. The notebook attached to the repo also serves as a tutorial for image classification with PyTorch.
Fruit Classification - (PyTorch)
Classification carried out on a dataset of over 100 fruits. The model is a custom model inheriting from nn.Module.
Fruits-360 Classification Repo
Text Generation - Character Level - (PyTorch)
An implementation of a text generation method with Pytorch, created with a recurrent neural network.
Text Generation with Pytorch RNN Repo
Neural Style Transfer - (PyTorch)
An example of neural style transfer done with PyTorch. The content of one image is redrawn with the style of another image by averaging the loss of the two images.
Transfer Results
Content Images
Style Images
Mushroom Classifier
Classification and visualization of a dataset consisting of mushrooms with many different features.
Video Game Sales
An analysis of video game sales and a comparison of several different sales prediction methods.
Video Game Sales Analysis and Prediction Repo
Sentiment Analysis
A custom sentiment analysis module was created, based on the voting of multiple classifiers like Naive Bayes, Logistic Regression, Decision Trees, and Stocastic Gradient Descent. The performance of this custom sentiment analysis module is compared with the performance of other sentiment analysis modules.
Titanic Analysis
An analysis and classfication done on the Titanic dataset. Data is preprocessed and missing values imputed, the data is then visualized and analyzed. Afterwards, multiple classifiers are run on the dataset and their performance is compared.
Data Analysis
Facebook Text Data Analysis
This project catalogues many functions I used to visualize and analyze my own Facebook data. The analysis was carried out on text data and a number of different analytical techniques - such as sentiment analysis, named entity recognition, and word clouds - were used.
GameInformer Data Analysis
Data was scraped from over 1100 reviews of games done by GameInformer. The data was then analyzed to explore patterns regarding review scores, publishers, developers, consoles and more.
GameInformer Review Analysis Repo
Other Python Projects
Movie Plot Generator
This project collected plot summaries from a variety of movies catalogued at the Open Movie Database. These plot summaries were then used to create new plot ideas for writers. The project compared multiple methods of generating text, including using an LSTM neural network, a Recurrent Neural Network and Markov Chains.
Movie Recommendation System
A movie recommendation system based on the similarity of movie attributes such as Director, cast, genre, and various keywords. The similarities between the movies are calculated with different metrics and averaged together.
Front End/Web Development
Javascript Ball Bounce Animation
An animation featuring multiple colored balls bouncing around, which gradually speed up.
Javascript Custom Media Database
A custom media database that allows the user to enter and track the media they engage with. They can enter the type of media, the completion status, and any notes they have.
Javascript Quiz App
A simple quiz app.
Javascript Weather App
A weather app that gets weather based on your current GPS location.
Django Blog App
A blog app created in Django. This app allows the creation and editing of blog posts by users that can create and edit their accounts.
Django Craigslist Wrapper
A wrapper for Craigslist that lets people make searches and view search results with an improved visual format.
Django Poll App
An app that allows users to see and vote on various polls.
Django To-Do App
A to-do app created in Django that lets the user create items on a to-do list and delete those items when finished.
Tutorials
Blood Cell Classifier - Transfer Learning - Model Finetuning (Keras)
This project demonstrates the use of transfer learning and finetuning to classify images of blood cells.
Blood Cell Classifier Repo and IPy Notebook
Keras Image Recognition Tutorial
Keras Image Recognition Tutorial Article
Multiple Linear Regression Tutorial
Multiple Linear Regression Tutorial Article
PyTorch Transfer Learning Tutorial
Transfer Learning PyTorch Tutorial Article
Data Analysis and Storage with MongoDB Tutorial
Data Analysis/MongoDB Tutorial Article
Overview of Classification Methods in Python/Scikit-Learn Tutorial
Scikit-learn Classification Tutorial Article