Making an enabled capability
View Patent ↗Various embodiments relate to network capabilities. Devices of a network can have different capabilities. The network can provide artificial intelligence (AI) enabled, machine learning (ML) enabled, deep learning (DL) enabled networked access to these capabilities. The capabilities can share a common AI/ML/DL-enabled open layer-based net-centric logical protocol architecture. Also, different features can be achieved through different layers. As an example, AI enabled access can be achieved through the application layer, ML enabled access and DL enabled access can be achieved through the presentation layer and the session layer, and network access is achieved through the transport layer, the network layer, the link layer, and the physical layer.
1 . A method, comprising:
receiving a sensor signal;
processing the sensor signal into a convertible form;
digitizing the sensor signal in the convertible form into an uncleared data set;
clearing the uncleared data set to produce a cleared data set;
performing feature extraction upon the cleared data set to produce a feature set through application of a dimensionality reduction algorithm;
performing feature classification upon the feature set to produce a classified feature set; and
causing output of the classified feature set.
2 . A method, comprising:
obtaining an uncleared data set;
clearing the uncleared data set to produce a cleared data set;
performing feature extraction upon the cleared data set to produce a feature set;
performing feature classification upon the feature set to produce a classified feature set; and
causing output of the classified feature set,
where performing feature extraction upon the cleared data set to produce the classified feature set comprises identifying an output application destination for the classified feature set, selecting a feature algorithm set based on the output application destination, applying the feature algorithm set on the cleared data set to produce a potential feature set, and applying a dimensionality reduction algorithm set on the potential feature set with the output of the dimensionality reduction algorithm set being the feature seta definition of noise that is based, at least in part, on the feature set andfiltering out the noise from the uncleared data set, the noise comprises a non-feature data entry.
3 . A method, comprising:
obtaining an uncleared data set;
clearing the uncleared data set to produce a cleared data set;
performing feature extraction upon the cleared data set to produce a feature set by identifying an output application destination for the classified feature set, selecting an algorithm set based on the output application destination, and applying the algorithm set on the cleared data set, the output of the algorithm set being the feature set, where the algorithm set is a dimensionality reduction algorithm set;
performing feature classification upon the feature set to produce a classified feature set; and
causing output of the classified feature set.
4 . The method of claim 1 , comprising:
employing the classified feature set, after being outputted, in a predictive test to produce a predictive test result; and
causing output of the predictive test result.
5 . The method of claim 4 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a machine learning algorithm upon the feature set.
6 . The method of claim 4 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a deep learning algorithm upon the feature set.
7 . The method of claim 2 , where obtaining the uncleared data set comprises:
receiving a sensor signal;
processing the sensor signal into a convertible form; and
digitizing the sensor signal in the convertible form into the uncleared data set.
8 . The method of claim 7 , comprising:
employing the classified feature set, after being outputted, in a predictive test to produce a predictive test result; and
causing output of the predictive test result.
9 . The method of claim 8 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a machine learning algorithm upon the feature set.
10 . The method of claim 8 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a deep learning algorithm upon the feature set.
11 . The method of claim 3 , where obtaining the uncleared data set comprises:
receiving a sensor signal;
processing the sensor signal into a convertible form; and
digitizing the sensor signal in the convertible form into the uncleared data set.
12 . The method of claim 8 , comprising:
employing the classified feature set, after being outputted, in a predictive test to produce a predictive test result; and
causing output of the predictive test result.
13 . The method of claim 9 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a machine learning algorithm upon the feature set.
14 . The method of claim 9 , where performing feature classification upon the feature set to produce a classified feature set comprises:
applying a deep learning algorithm upon the feature set.
15 . The method of claim 3 , comprising:
employing the classified feature set, after being outputted, in a predictive test to produce a predictive test result; and
causing output of the predictive test result.
16 . The method of claim 2 , comprising:
employing the classified feature set, after being outputted, in a predictive test to produce a predictive test result; and
causing output of the predictive test result.