Scalable multiple layer machine learning model for classification of Wi-Fi issues on a data communication network
AssigneeFortinet, Inc.
InventorsSiva Yogendra Jupudi, Deepti Girish, Shunmugaraj Karuvanayagam
Life of this patent
Estimating the term…
Abstract
Multi-level machine learning models can be generated from the captured log events. Outcomes are predicted for input events in real-time. The captured log events are received and parsed to expose event outcome data. A first data set is generated by determining whether an outcome associated with the event outcome data was a success or a failure. Responsive to a failed event outcome, a second data set is generated by categorizing the failed event outcome, to train multiple level SVMs for prediction of Wi-Fi input events and automatic remediation of Wi-Fi issues.
Claims12
A network device coupled to receive Wi-Fi event logs of a private data network, for classifying Wi-Fi log events of a data communication system in multiple layers using machine learning, the network device comprising:
- a processor;
- a network interface communicatively coupled to the processor and to the private data network;
- and
- a memory, storing:
- an event pre-processor module to receive Wi-Fi log events and parse the received Wi-Fi log events to expose event outcome data, wherein the log events are string match operations;
- an event classifier module to generate a first data set by determining whether an outcome associated with the event outcome data was a success or a failure, and responsive to a failed event outcome, to generate a second data set by categorizing the failed event outcome;
- a machine learning engine to train multiple level support vector machines (SVMs), wherein a first model of SVMs is trained using the first data set and a second model of SVM is trained using the second data set;
- and
- an input prediction engine to predict an outcome of a Wi-Fi input event in real-time based on the multiple level SVMs.
A method in a network device coupled to receive Wi-Fi event logs of a private data network, for classifying Wi-Fi log events of a data communication system in multiple layers using machine learning, the method comprising the steps of:
- receiving Wi-Fi log events and parsing the received Wi-Fi log events to expose event outcome data, wherein the log events are string match operations;
- generating a first data set by determining whether an outcome associated with the event outcome data was a success or a failure;
- responsive to a failed event outcome, generating a second data set by categorizing the failed event outcome;
- training multiple level support vector machines (SVMs), wherein a first model of SVMs is trained using the first data set and a second model of SVM is trained using the second data set;
- and
- predicting an outcome of a Wi-Fi input event in real-time based on the multiple level SVMs.
A non-transitory computer-readable media in a network device coupled to receive Wi-Fi event logs of a private data network for, when executed by a processor, for classifying Wi-Fi log events of a data communication system in multiple layers using machine learning, the method comprising the steps of:
- receiving Wi-Fi log events and parsing the received Wi-Fi log events to expose event outcome data, wherein the log events are string match operations;
- generating a first data set by determining whether an outcome associated with the event outcome data was a success or a failure;
- responsive to a failed event outcome, generating a second data set by categorizing the failed event outcome;
- training multiple level support vector machines (SVMs), wherein a first model of SVMs is trained using the first data set and a second model of SVM is trained using the second data set;
- and
- predicting an outcome of a Wi-Fi input event in real-time based on the multiple level SVMs.
Dependencies are inferred from the claim text.