Dockerized application to predict stock market fluctuation. Uses datamining techniques to form a kNN model based on processed social media data using NLP, emotional data from faces in videos using OpenCV and Tesseract, and bullish/bearish values using DT to predict stock market trends.
RouteBot is an iOS application I made in the Summer of 2017 to introduce myself to the Swift Programming Language. It helps runners by calculating possible routes based on a location and desired mileage. My app incorporates an algorithm I designed to create well-formed routes for users, which involves manipulating coordinates, recursive event handling, and outlier removal.
Surveillance Footage Blurring
Surveillance Footage Blurring is a project that addresses concerns brought up in Security and Social Dimensions of City Surveillance Policy. The paper mentions a need for city surveillance systems to anonymize their captured for public consumption. The solution proposed in the 2014 paper suggested a manual solution where you have government worker using video editing software to blur the myriad of footage obtained. My project automates this task by taking a video file and altering frame by frame. I am using OpenCV and OpenALPR for object detection, where I gather the coordinates for a detected face or license plate, and then use a Gaussian Blurring technique on the space between those coordinates. The result is a video with all personal data eliminated. I am currently working on gathering more surveillance footage so I can train my own Haar Cascade based on that footage for more accurate detection.
Mountain Meter is a rapid prototype for my Mobile Development Course project. It gathers sensor data from Apple's CoreMotion Library and contextualizes the data in a way that benefits active users. The app gathers data about the users activity (flights of stairs ascended, distance travelled) and compares it to real world activities. So for example, the user can compare the flights of stairs they've ascended in the past week to the height of Mount Everest.
Beta Code Notification is a weekend project I created using Python and the Twitter API to notify followers when a new beta program becomes available. I always found myself missing out on the oppurtunities, and never found a reliable way to stay informed. The project scrapes a well known video game website that has a beta test news section, my bot parses the feed, stores new info in a SQLite3 database, and another script comes along and posts the new betas to the twitter account. Tweets by BetaCodeNotif
Amy bot is a reddit bot that uses the Reddit API to gather the stream of comments on the Reddit site. It parses all content searching for amazon links posted by the sites users, and then replies back to the users with a shortened version of the URL and also a Amazon Smile version of the link as well. The script also records some meta data about the user, and particular section of the site the link was posted to into an sqlite3 database.
Schedule Checker is a Python script that will notify a user via email that an FSU class is available. The project uses the Selenium framework to log on to the site and navigate the page, and uses beautifulsoup4 to scrape the html, and then parses the data. This project stemmed from my frustrations of having to continuously check if a Python class had an available seat open. It can now be used by any student to check if any course is available.
Draftr was a project for my Software Engineering course in Fall 2015. It's designed to be a tool for casual fantasy football players to easily and quickly track football stats from the past five years. The website used Python along with the Django framework to communicate with the PostgreSQL database(provided by the open source project nfldb) to retrieve live statistics from games across the country.
During my internship at Marquis Software, they posed to all of their employees the question:
What is the probability that at least 6 out of 60 people have a birthday in the same 10 day period?I solved this with a simple Java program that calculated the correct answer, and as a result won the challenge.