System and method for detecting generalized space-time clusters

A method for detecting clusters in space and time using input data on occurrences of a disease in a population and characteristics of the population at a plurality of geographic locations and times comprising:

• determining expected occurrences of the disease at a plurality of geographic locations and a plurality of times;
• determining actual occurrences of the disease at a plurality of geographic locations and a plurality of times;
• searching the expected occurrences and the actual occurrences for a plurality of candidate solutions, wherein each solution is represented as a set of points in a three-dimensional space, and wherein each point corresponds to a geographic location at a time;
• determining at least one solution corresponding to a selected convex container shape from the plurality of candidate solutions, wherein the convex container module determines a solution that maximizes a measure representing the equality between the set of points in the given candidate solution and the set of points included in the solution, given the selected convex container shape;
• and
• determining a strength metric for each solution determined by the convex container module, the search module selecting a solution having a desirable strength, wherein the solution having the desirable strength indicates a dominant cluster in the input data corresponding to a detected cluster of the disease in the population.