Mark Twain once advised that we should not allow our schooling to interfere with our education. While I agree with this sentiment, I firmly believe that organized education is nonetheless important in providing a developing mind with the structure and discipline it often requires to mature. Ergo, my education and my schooling are the same: where one system is deficient, the other method is employed as a supplement.
To the best of my ability, I endeavor to complete projects that are both of interest and value to my education. Below you will find a sample of polished and curated works relating to the computer science and engineering disciplines. I have segregated these projects based on their origin but please take note that ideas immigrate and a project that once began in school as research can often find itself crossing the perilous chasm to the new and exciting domains of application and business.
Statistically analyzed proprietary municipal water quality data for a private chemical company. Researched statistical and machine learning solutions for predicting water degradation.
Compared the predictive power of long short-term memory (LSTM) artificial neural networks (ANNs) against the standard convolutional neural network (CNN) approach to time-series forecasting.GitHub
Replicated Avery Wang's industrial-strength audio search algorithm (used in Shazam and SoundHound). Tested the speed at which the algorithm was able to positively identify a recognized song based on a pre-specified confidence interval.GitHub Report
Wrote a C program to sample an accelerometer and gyroscope breakout board (LSM303) from a Beagle Bone microcomputer through the I2C bus, allowing it to track it's position and velocity through space. The movement history of the device was then graphed using Python.GitHub Report
A contemporary and comprehensive literature review of 5G small-cell antenna technology and how cities safely plan and implement their installation on transportation infrastructure.Blog Publication
An investigation into the use of inexpensive RFID tags to simultaneously localize and map (SLAM) a robots environment with high-fidelity.
Screen scrapes a table of upcoming holidays and dates and serves the data RESTfully through a Flask API. This project was successfully completed as part of an interview at Nerevu Group.GitHub
Wraps nvidia-smi to actively monitor the status of GPU components and automatically email the administrator to warn them when the GPU exceeds the limitations prescribed for it (e.g. temperature).GitHub
A study of the metabolic equivalent for task (MET) based calculation behind estimates of caloric expenditures during cycling events. Includes both the discrete MET (DMET) calculation and the continuous MET (CMET) calculation developed through machine learning methods.Blog GitHub
A crude analysis of precious metals commodities vis a vis their relative risk level against the market. The data wrangling that occurs therein can be expanded upon to answer more complex questions about the nature of the commodities market.Blog GitHub
Demonstration of authentication procedures using Yubico OTP keyfobs.GitHub