Event identification in sensor analytics

A non-transitory computer readable medium having stored thereon a computer program that, when executed, causes a computer to perform the steps of:

• receiving one or more individualized data points, wherein each individualized data point represents a spatial or temporal event and has a unique identifier associated with it;
• distributing the one or more individualized data points across a grid, wherein the grid includes one or more cells;
• determining an event likelihood ratio for each of one or more of the grid cells, wherein determining the event likelihood ratio for a respective grid cell comprises determining an expected number of individualized data points in the respective grid cell;
• identifying one or more event clusters, based on the event likelihood ratio, wherein each event cluster includes one or more of the grid cells, thus defining a region on the grid;
• generating multiple replicas of the region;
• performing simulations on the multiple replicas to obtain a significance value for each of the one or more event clusters;
• and
• storing the one or more event clusters in a data repository, based on a comparison of the significance value to a threshold.

A system for detecting an event anomaly comprising:

• a network;
• one or more user devices in communication with the network;
• and
• a processor in communication with the network, wherein the processor is for:
• i) receiving one or more individualized data points representing data originating from the one or more user devices, wherein each individualized data point represents a spatial or temporal event and has a unique identifier associated with it;
• ii) distributing the one or more individualized data points across a grid, wherein the grid includes one or more cells;
• iii) determining an event likelihood ratio for each of one or more of the grid cells, wherein determining the event likelihood ratio for a respective grid cell comprises determining an expected number of individualized data points in the respective grid cell;
• iv) identifying one or more event clusters, based on the event likelihood ratio, wherein each event cluster includes one or more of the grid cells, thus defining a region on the grid;
• v) generating multiple replicas of the region;
• vi) performing simulations on the multiple replicas to obtain a significance value for each of the one or more event clusters;
• and
• vii) storing the one or more event clusters in a data repository, based on a comparison of the significance value to a specified threshold value.

A system comprising:

• means for distributing one or more individualized data points across a grid, wherein the grid includes one or more cells;
• means for detecting an event likelihood ratio for each of the one or more cells of the grid, wherein detecting the event likelihood ratio for a respective grid cell comprises determining an expected number of individualized data points in the one or more cells;
• means for identifying one or more event cluster based on the event likelihood ratio, wherein each event cluster includes one or more of the grid cells, thus defining a region on the grid;
• means for generating multiple replicas of the region;
• means for performing simulations on the multiple replicas to obtain a significance value for each of the one or more event cluster;
• and
• means for storing the one or more event clusters in a data repository, based on a comparison of the significance value to a threshold.

A system comprising:

• means for distributing one or more individualized data points across a grid, wherein the grid includes one or more cells;
• means for detecting an event likelihood ratio for each of the one or more cells of the grid, wherein detecting the event likelihood ratio for a respective grid cell comprises determining an expected number of individualized data points in the one or more cells;
• means for identifying one or more event cluster based on the event likelihood ratio, wherein each event cluster includes one or more of the grid cells, thus defining a region on the grid;
• means for generating multiple replicas of the region;
• means for performing simulations on the multiple replicas to obtain a significance value for each of the one or more event cluster, wherein the significance value corresponds to a ratio of the number of simulations that result in a maximum significance score that is greater than the event likelihood ratio relative to a total number of simulations performed;
• and
• means for storing the one or more event clusters in a data repository when the significance value is determined to be less than a threshold risk level.