Application of biologically inspired neural oscillators to colour image segmentation

Developing Architectures of Spiking Neural Networks by Using Grammatical Evolution Based on Evolutionary Strategy

“…When a SNN is applied to solve pattern recognition problems, original patterns can not be fed into the SNN; they need to be transformed as spikes in an interval of time by using an encoding scheme. Several encoding schemes have been proposed, such as the Gaussian Receptive Fiels (GRFs); this encoding scheme has been extensively used in [2], [1], [21]. Basically, this encoding scheme requires for encoding a variable, m neurons (with Gaussian functions) used for covering the whole range of the variable, γ as a coefficient for setting the width of Gaussian functions and the encoding simulation time τ .…”

“…The architecture of the SNN can vary depending on the kind of the problem to solve. This work focuses on SNNs with architecture known as Fully-Connected Feed-Forward as used in [2], [1] and [11]. There are several spiking neuron models that can be implemented in a SNN, this work use the Spike Response Model which is detailed in next section.…”

“…For this work, the spiking neurons use the time-tofirst-spike as coding scheme for sending/receiving messages. Due to the coding scheme being used, a reduced version of the SRM is implemented, which has been used in [2], [1] and [21].…”

Structured Memetic Automation for Online Human-Like Social Behavior Learning

“…At the second phase (careful search), meme activation, meme competition and meme matching are used to find the best matching meme from the best and smallest active subclass (lines 7-9). At the third phase (memotype update), the memetic agent updates its structured memes based on the matching result (lines [10][11][12][13][14][15][16]. If the current vector }S, A, R| matches an existing meme successfully, memotype update is done for the best matching meme.…”

“…Since the number of meme blocks grows significantly in a complex problem domain, the meme search becomes inefficient on providing decision support. Following the line of biologically inspired representation [16], we propose memetic agents with structured memes to speed up decision making of the memetic automatons. The representation implies that humans learn a complex concept by first identifying simple and abstract ones and then composing them together.…”