Predicting a Happier Place

In communities with higher well-being, people live longer and more productive lives, and local economies flourish. In an effort to improve understanding of what leads to vibrant communities and fulfilling lifestyles, our project explored community-level factors and their significance for determining well-being in Americas largest metropolitan areas. We hypothesized that the frequency of points of interests (POIs) in a community as well as basic macro-level factors were sufficiently accurate indicators of well-being. Thus, the input to our algorithm was POI frequencies and economic/social factors for each metro area. We then used Logistic Regression, Extra-Trees Classifier, and Support Vector Machines to output a prediction of whether a metro area is of high or low wellbeing.