Electromagnetic noise signal based predictive analytics

A method for predicting user touch events, the method comprising:

• receiving, by one or more computer processors, a detected electromagnetic noise signal of a first object, wherein the electromagnetic noise signal of the first object is detected by a smart watch that is conductively connected to a wearer of the smart watch;
• comparing, by the one or more computer processors, the detected electromagnetic noise signal of the first object to one or more stored electromagnetic noise signals associated with one or more objects;
• based, at least in part, on the comparison and metadata that identifies a name and a location of the first object, determining, by the one or more computer processors, the identity of the first object as a kitchen appliance;
• responsive to determining the identity of the first object as a kitchen appliance, storing, by the one or more computer processors, metadata corresponding to an electromagnetic noise signal detection event associated with the kitchen appliance;
• determining, by the one or more computer processors, whether a first quantity and a first frequency of recorded metadata corresponding to the electromagnetic signal detection event associated with the kitchen appliance meets a learning threshold;
• responsive to determining that the first quantity and the first frequency of the recorded metadata corresponding to the electromagnetic signal detection event associated with the kitchen appliance meets the learning threshold, predicting, by the one or more computer processors, a first subsequent electromagnetic noise signal detection event associated with the kitchen appliance (i) based, at least in part, on one or more time stamps of touch events associated with the kitchen appliance, a frequency of touch events associated with the kitchen appliance, and one or more categories of objects corresponding to touch events before and after the touch events associated with the kitchen appliance, and (ii) utilizing a predictive analytics algorithm selected from the group consisting of a time series model and a supervised learning classifier, and wherein the supervised learning classifier is a regression analysis;
• based, at least in part, on a time of the predicted subsequent electromagnetic noise signal detection event associated with the kitchen appliance, determining, by the one or more computer processors, a range of times at which to display a produce advertisement that is associated with a meal of the wearer of the smart watch;
• displaying, by the one or more computer processors, the produce advertisement at a time within the range of times on one or more computing devices associated with the wearer of the smart watch;
• receiving, by one or more computer processors, a detected electromagnetic noise signal of a second object;
• comparing, by the one or more computer processors, the detected electromagnetic noise signal of the second object to the one or more stored electromagnetic noise signals associated with one or more objects;
• in response to determining, by the one or more computer processors, that the detected electromagnetic noise signal of the second object cannot be identified among the one or more stored electromagnetic noise signals, prompting, by the one or more computer processors, the user to input metadata including a type of object of the second object and a location of second object;
• determining, by the one or more computer processors, whether a second quantity and a second frequency of recorded metadata corresponding to the electromagnetic signal detection event associated with the second object meets the learning threshold;
• responsive to determining that the second quantity and the second frequency of the recorded metadata corresponding to the electromagnetic signal detection event associated with the second object does not meet the learning threshold, receiving, by the one or more computer processors, one or more additional detected electromagnetic noise signals associated with the second object;
• and
• storing, by the one or more computer processors, additional metadata associated with the one or more additional detected electromagnetic noise signals.