AZIONE DI ACQUISTO BITCOIN

I first create a high-level Twitter Anonymity Dataset, based off user's level of anonymity on the platform. Then I explore the communicative effects that users demonstrate in aggregate based on their level of identifiability online. Successfully defended Honors Thesis that was highlighted at the 2021 BYU Honors Conference.

Recreated popular machine learning models in Python to gain a deeper understanding of their workings. Including, decision tree, random forest, adaboost, linear regression, perceptron, support vector machines, etc.

A Monte Carlo Tree Search implementation of an agent for the board game, Ticket to Ride. Based off 24-city European edition, with 50 routes. The agent is able to consistently beat average players.

Utilizing the Twitter API, the model constructs a novel, real-time dataset based on the run-time. Then it preprocesses the data, classifies each tweet by a fitted emotion classifier, creates an inverted index that users can then quickly query.

For this project, we are attempting to properly predict whether a person makes over $50,000 annually based on a number of features. We explore the efficacy of various models, explore feature engineering, and fit a Random Forest Classifier that is able classify 91% of datapoints without overfitting. This was top 5 in a graduate ML course of 100 students

Exploration of different approaches to tackling the TSP problem, a difficult route-finding problem that is classified as NP-Complete. I first make a greedy solution as a baseline, and then implement a branch-and-bounding solution.

Attempting to predict the likelihood of a person defaulting on their credit based off a variety of features.