Decision making based on spike time dependent plasticity of a neural network

A system comprising:

• one or more processors;
• and
• memory for storing instructions that, when executed by the one or more processors, cause the system to:
• classify inputs using input layer-to-hidden layer connections of a neural network model;
• based on the input classifications, make decisions using hidden layer-to-output layer connections of the neural network model;
• responsive to the input classifying, modulate the input layer-to-hidden layer connections using non-rewarded spike time dependent plasticity (STDP) signals;
• responsive to the decision making, modulate the hidden layer-to-output layer connections using rewarded STDP signals;
• and
• responsive to the modulation of the input layer-to-hidden layer connections and hidden layer-to-output layer connections, reiterate the input classifying of the input layer-to-hidden layer connections and the decision making of the hidden layer-to-output layer connections.

Non-transitory computer-readable medium storing instructions, which when executed by one or more processing resources, cause the one or more processing resources to:

• classify inputs using input layer-to-hidden layer connections of a neural network model;
• based on the input classifications, make decisions using hidden layer-to-output layer connections of the neural network model;
• responsive to the input classifying, modulate the input layer-to-hidden layer connections using non-rewarded spike time dependent plasticity (STDP) signals;
• responsive to the decision making, modulate the hidden layer-to-output layer connections using rewarded STDP signals;
• and
• responsive to the modulation of the input layer-to-hidden layer connections and hidden layer-to-output layer connections, reiterate the input classifying of the input layer-to-hidden layer connections and the decision making of the hidden layer-to-output layer connections;
• wherein:
• the neural network model further comprises an input layer, a plurality of hidden layers, and an output layer, and
• the input layer-to-hidden layer connections connect the input layer to a first hidden layer of the plurality of hidden layers and the hidden layer-to-output layer connections connect a second hidden layer of the plurality of hidden layers to the output layer.