Network reconfiguration using genetic algorithm-based predictive models

A device comprising:

• a processing system including at least one processor;
• and
• a computer-readable medium storing instructions which, when executed by the processing system, cause the processing system to perform operations, the operations comprising:
• gathering training data comprising a second plurality of measurements of a plurality of performance indicators and a plurality of measurements of a target performance indicator;
• determining a final weight set comprising weight factors to apply to each of the plurality of performance indicators for a predictive model associated with the target performance indicator using a genetic algorithm, wherein the genetic algorithm comprises:
• generating a first plurality of candidate weight sets for a first generation;
• determining fitness parameters for the first plurality of candidate weight sets, wherein each of the fitness parameters comprises a measure of a correlation between an output of the predictive model using a candidate weight set of the first plurality of candidate weight sets and a measure of the target performance indicator from the training data;
• selecting weight set pairs from the first plurality of candidate weight sets for generating a second plurality of candidate weight sets in a second generation, wherein each of the weight set pairs is selected from the first plurality of candidate weight sets probabilistically based upon the fitness parameters of the first plurality of candidate weight sets;
• and
• generating the second plurality of candidate weight sets in the second generation from the weight set pairs, wherein the generating the second plurality of candidate weight sets in the second generation from the weight set pairs comprises, for a weight factor of a candidate weight set of the second plurality of candidate weight sets, determining whether to apply a mutation, wherein the determining whether to apply the mutation is in accordance with a mutation ratio, wherein the mutation comprises a value randomly selected between a maximum value and a minimum value for the weight factor;
• gathering a first plurality of measurements of the plurality of performance indicators for at least a portion of a cellular network;
• applying the predictive model to the first plurality of measurements of the plurality of performance indicators to generate a prediction for the target performance indicator;
• and
• adjusting at least one aspect of the cellular network in response to the prediction.

A non-transitory computer-readable medium storing instructions which, when executed by a processing system including at least one processor, cause the processing system to perform operations, the operations comprising:

• gathering training data comprising a second plurality of measurements of a plurality of performance indicators and a plurality of measurements of a target performance indicator;
• determining a final weight set comprising weight factors to apply to each of the plurality of performance indicators for a predictive model associated with the target performance indicator using a genetic algorithm, wherein the genetic algorithm comprises:
• generating a first plurality of candidate weight sets for a first generation;
• determining fitness parameters for the first plurality of candidate weight sets, wherein each of the fitness parameters comprises a measure of a correlation between an output of the predictive model using a candidate weight set of the first plurality of candidate weight sets and a measure of the target performance indicator from the training data;
• selecting weight set pairs from the first plurality of candidate weight sets for generating a second plurality of candidate weight sets in a second generation, wherein each of the weight set pairs is selected from the first plurality of candidate weight sets probabilistically based upon the fitness parameters of the first plurality of candidate weight sets;
• and
• generating the second plurality of candidate weight sets in the second generation from the weight set pairs, wherein the generating the second plurality of candidate weight sets in the second generation from the weight set pairs comprises, for a weight factor of a candidate weight set of the second plurality of candidate weight sets, determining whether to apply a mutation, wherein the determining whether to apply the mutation is in accordance with a mutation ratio, wherein the mutation comprises a value randomly selected between a maximum value and a minimum value for the weight factor;
• gathering a first plurality of measurements of the plurality of performance indicators for at least a portion of a cellular network;
• applying the predictive model to the first plurality of measurements of the plurality of performance indicators to generate a prediction for the target performance indicator;
• and
• adjusting at least one aspect of the cellular network in response to the prediction.