Luis Ulloa

Housing Price Prediction

Problem Statement: In the United States people value spacious properties. Therefore, it is assumed that the more living space is available in a house, and the bigger the lot, the higher the price the property would command in the local market. This analysis aims to build a predictive pricing model based on that premise. The model will utilize multiple linear regression techniques, and the model's performance will be assessed using the metric Root Mean Squared Error (RMSE). Once the model is built and refined, it will be used to predict sales prices as part of an internal Kaggle challenge.

Link to project repo: GitHub

Exit

Reddit Post Classification

Problem Statement: How smart can a classification algorithm be? The human mind can understand and process written language quite efficiently. If given a random blurb with enough words and context, a person can easily guess what the topic is about with a high degree of accuracy. Can a classification model mimic such prowess? This project aims to answer that question. In order to do so, different classification models will be built and trained to read randomized text from two different subreddits and then predict what topic is being discussed. Easy for a person, can the machine beat the person? Various metrics will be used to evaluate each model's performance, with a key focus on misclassification. The winning model shall be the one with the lowest misclassification rate.

Link to project repo: GitHub

Exit

Ride BigApple

Problem Statement: Can yellow cab fares be predicted within New York City's five boroughs based on time of the day, time of the year and certain high passenger areas? The aim here is to use existing historical data of NYC cab rides spanning several years and building a model that can be integrated into a simple app to request a taxi ride and predict the expected fare and distance for the passenger.

Link to project repo: GitHub

Exit