This portfolio is pretty dated and reflects the work I did when I was trying to become a Data Scientist - for a more recent look at my work go to: wwww.christophershoe.com.
- Project 1: SAT Scores + Summary Statistics
- Project 2: Billboard Hits + Data Munging
- Project 3: Liquor Sales + Linear Regression
- Project 4: Web Scraping + Logistic Regression
- Project 5: Disaster Relief + Classification
- Project 6: IMDB API + Random Forests
- Project 7: Airport Delays + Cluster Analysis
During my tenure at General Assembly I completed 7 projects that served to showcase my skill sets and understanding of machine learning, statistics, python, and practical application of Data Science methodologies to real world problems. The readme files serve as a primer to understanding the code and approach taken. All solutions can be found in the corresponding project's "solutions" folder - labeled as that project's name.
Skills Showcased: Python | SQL | Tableau | Machine Learning | Web Scraping | Regex | Pandas | NumPy | Linear & Logistic Regression modeling | Random Forests | NLP | AWS | Technical Writing | Object Oriented Programing
Building off the TensorFlow model magenta, Project "Miles Ahead" attempts to expand Deep Learning Jazz Generation by enhancing the recurrent neural network "Lookback RNN".
In addition, through the use of Music21 and matplotlib, Project Miles Ahead will conduct Exploratory Data Analysis and robust Data Visualization to explore the music created by the model. Utilizing the same comparison on music created by legendary jazz pianist, Bill Evans, this project seeks insight in to the "humanity" of the computer generated music.
The Astroid Grand Challenge is a large scale effort together by NASA, utilizing multiple industries and disciplines, to better detect and predict astroid threats in our solar system and protect the Earth from a wide-scale tragedy.
From NASA's 2016 Space App Challenge:
"There are millions of yet undiscovered Near Earth Objects (NEOs) which could pose a threat to Planet Earth. These Asteroids require space-based hardware to locate and track, however once their position is identified, follow-up observations can be made with radar or optical telescopes gathering light curve data - enabling estimates of composition, reflectivity, rotation and other characteristics that inform mitigation strategies to deflect objects before they impact with Earth. Presently, only a handful of hazardous NEOs have been detected prior to entering our atmosphere. The immense task of asteroid hunting is further complicated by the high number of false positives and long duration between observations - where some NEOs have orbits of many decades. Presented with these challenges, the space community has begun to look towards "machine learning" to both mechanize and accelerate the speed of detection and characterization."
Problem Statement Summary
Utilizing provided data from the Minor Planet Center, and NASA's NEOWISE, is it possible to build a predictive model - based on machine learning algorithms - that can accurately detect and characterize Near Earth Objects on a trajectory for Earth? If successful, such a model could be sufficient to add to the body of work found in the Astroid Grand Challenge and further impact our ability to protect our Earth from sharing the same fate as the dinosaurs.
Project is in progress and will be updated as work is completed.