Searching for a new home or storefront location in Baltimore, but don't know where to begin? Let our data be your guide!
Select and weight the factors you care about by checking either the negative () or positive () boxes below, then click submit to view recommended hotspots.
In order to generate a recommendation, you must provide at least one positive factor.
We begin by collecting raw datasets from a variety of sources such as Socrata, Yelp, and the National Parks Service.
Individual datasets can be visualized on a map as shown to the right (below, on mobile), but plotting multiple maps at once doesn't yield much visual insight.
Rather than plotting all of the points in a dataset, we can visualize each dataset's density using a technique known as kernel density estimation.
When a user enters their preferences and clicks 'submit', the server computes the sum of the weighted density maps and returns latitude and longitude coordinates corresponding to areas with the highest overall density.
We're working on applying unsupervised machine learning algorithms, such as K-means clustering, in order to uncover hidden neighborhood dynamics and relationships. Check out this page to see some of the progress we've made.