Limits...
Multiclass Classification by Adaptive Network of Dendritic Neurons with Binary Synapses Using Structural Plasticity.

Hussain S, Basu A - Front Neurosci (2016)

Bottom Line: The performance of the model is evaluated on classification of handwritten digits from the benchmark MNIST dataset and compared with other spike classifiers.For VLSI implementations, we show that the reduced synaptic memory can save upto 4X area compared to conventional crossbar topologies.Finally, we also present a biologically realistic spike-based version for calculating the correlations required by the structural learning rule and demonstrate the correspondence between the rate-based and spike-based methods of learning.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical and Electronic Engineering, Nanyang Technological University Singapore, Singapore.

ABSTRACT
The development of power-efficient neuromorphic devices presents the challenge of designing spike pattern classification algorithms which can be implemented on low-precision hardware and can also achieve state-of-the-art performance. In our pursuit of meeting this challenge, we present a pattern classification model which uses a sparse connection matrix and exploits the mechanism of nonlinear dendritic processing to achieve high classification accuracy. A rate-based structural learning rule for multiclass classification is proposed which modifies a connectivity matrix of binary synaptic connections by choosing the best "k" out of "d" inputs to make connections on every dendritic branch (k < < d). Because learning only modifies connectivity, the model is well suited for implementation in neuromorphic systems using address-event representation (AER). We develop an ensemble method which combines several dendritic classifiers to achieve enhanced generalization over individual classifiers. We have two major findings: (1) Our results demonstrate that an ensemble created with classifiers comprising moderate number of dendrites performs better than both ensembles of perceptrons and of complex dendritic trees. (2) In order to determine the moderate number of dendrites required for a specific classification problem, a two-step solution is proposed. First, an adaptive approach is proposed which scales the relative size of the dendritic trees of neurons for each class. It works by progressively adding dendrites with fixed number of synapses to the network, thereby allocating synaptic resources as per the complexity of the given problem. As a second step, theoretical capacity calculations are used to convert each neuronal dendritic tree to its optimal topology where dendrites of each class are assigned different number of synapses. The performance of the model is evaluated on classification of handwritten digits from the benchmark MNIST dataset and compared with other spike classifiers. We show that our system can achieve classification accuracy within 1 - 2% of other reported spike-based classifiers while using much less synaptic resources (only 7%) compared to that used by other methods. Further, an ensemble classifier created with adaptively learned sizes can attain accuracy of 96.4% which is at par with the best reported performance of spike-based classifiers. Moreover, the proposed method achieves this by using about 20% of the synapses used by other spike algorithms. We also present results of applying our algorithm to classify the MNIST-DVS dataset collected from a real spike-based image sensor and show results comparable to the best reported ones (88.1% accuracy). For VLSI implementations, we show that the reduced synaptic memory can save upto 4X area compared to conventional crossbar topologies. Finally, we also present a biologically realistic spike-based version for calculating the correlations required by the structural learning rule and demonstrate the correspondence between the rate-based and spike-based methods of learning.

No MeSH data available.


Ensemble classifier combining the intermediate class-specific outputs  to compute the combined class outputs Oμ, which are compared by the WTA to generate the final classifier output .
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4814530&req=5

Figure 2: Ensemble classifier combining the intermediate class-specific outputs to compute the combined class outputs Oμ, which are compared by the WTA to generate the final classifier output .

Mentions: We have also used an ensemble method in this work, where several classifiers when combined together yield better classification accuracy than any of the single classifiers in the ensemble. Our ensemble model consists of individually trained NLD classifiers which are then combined to classify novel test patterns. Since, previous research has demonstrated that combining identical classifiers doesn't produce any gain over individual classifiers (Opitz and Maclin, 1999), hence we created ensemble by combining several classifiers which are individually trained and disagree on their predictions. The complementary information about the novel patterns obtained from different classifiers can be exploited to produce a more accurate overall output. In order to generate different predictions for different classifiers, individual networks were initialized with different random synaptic connections. Figure 2 shows the basic framework for the classifier ensemble scheme. It consists of N individually trained multiclass classifiers as members of the ensemble. Each classifier generates the intermediate output, , which is the difference in outputs of PDT and NDT for class μ of the nth classifier. The intermediate outputs are combined in a class-specific manner to give . Finally, the ensemble output is generated using:(19)y^μ(x)=g(Oμ(x)−Oν(x)),Oν(x)≥Oξ(x),∀ν,ξ=μ


Multiclass Classification by Adaptive Network of Dendritic Neurons with Binary Synapses Using Structural Plasticity.

Hussain S, Basu A - Front Neurosci (2016)

Ensemble classifier combining the intermediate class-specific outputs  to compute the combined class outputs Oμ, which are compared by the WTA to generate the final classifier output .
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4814530&req=5

Figure 2: Ensemble classifier combining the intermediate class-specific outputs to compute the combined class outputs Oμ, which are compared by the WTA to generate the final classifier output .
Mentions: We have also used an ensemble method in this work, where several classifiers when combined together yield better classification accuracy than any of the single classifiers in the ensemble. Our ensemble model consists of individually trained NLD classifiers which are then combined to classify novel test patterns. Since, previous research has demonstrated that combining identical classifiers doesn't produce any gain over individual classifiers (Opitz and Maclin, 1999), hence we created ensemble by combining several classifiers which are individually trained and disagree on their predictions. The complementary information about the novel patterns obtained from different classifiers can be exploited to produce a more accurate overall output. In order to generate different predictions for different classifiers, individual networks were initialized with different random synaptic connections. Figure 2 shows the basic framework for the classifier ensemble scheme. It consists of N individually trained multiclass classifiers as members of the ensemble. Each classifier generates the intermediate output, , which is the difference in outputs of PDT and NDT for class μ of the nth classifier. The intermediate outputs are combined in a class-specific manner to give . Finally, the ensemble output is generated using:(19)y^μ(x)=g(Oμ(x)−Oν(x)),Oν(x)≥Oξ(x),∀ν,ξ=μ

Bottom Line: The performance of the model is evaluated on classification of handwritten digits from the benchmark MNIST dataset and compared with other spike classifiers.For VLSI implementations, we show that the reduced synaptic memory can save upto 4X area compared to conventional crossbar topologies.Finally, we also present a biologically realistic spike-based version for calculating the correlations required by the structural learning rule and demonstrate the correspondence between the rate-based and spike-based methods of learning.

View Article: PubMed Central - PubMed

Affiliation: School of Electrical and Electronic Engineering, Nanyang Technological University Singapore, Singapore.

ABSTRACT
The development of power-efficient neuromorphic devices presents the challenge of designing spike pattern classification algorithms which can be implemented on low-precision hardware and can also achieve state-of-the-art performance. In our pursuit of meeting this challenge, we present a pattern classification model which uses a sparse connection matrix and exploits the mechanism of nonlinear dendritic processing to achieve high classification accuracy. A rate-based structural learning rule for multiclass classification is proposed which modifies a connectivity matrix of binary synaptic connections by choosing the best "k" out of "d" inputs to make connections on every dendritic branch (k < < d). Because learning only modifies connectivity, the model is well suited for implementation in neuromorphic systems using address-event representation (AER). We develop an ensemble method which combines several dendritic classifiers to achieve enhanced generalization over individual classifiers. We have two major findings: (1) Our results demonstrate that an ensemble created with classifiers comprising moderate number of dendrites performs better than both ensembles of perceptrons and of complex dendritic trees. (2) In order to determine the moderate number of dendrites required for a specific classification problem, a two-step solution is proposed. First, an adaptive approach is proposed which scales the relative size of the dendritic trees of neurons for each class. It works by progressively adding dendrites with fixed number of synapses to the network, thereby allocating synaptic resources as per the complexity of the given problem. As a second step, theoretical capacity calculations are used to convert each neuronal dendritic tree to its optimal topology where dendrites of each class are assigned different number of synapses. The performance of the model is evaluated on classification of handwritten digits from the benchmark MNIST dataset and compared with other spike classifiers. We show that our system can achieve classification accuracy within 1 - 2% of other reported spike-based classifiers while using much less synaptic resources (only 7%) compared to that used by other methods. Further, an ensemble classifier created with adaptively learned sizes can attain accuracy of 96.4% which is at par with the best reported performance of spike-based classifiers. Moreover, the proposed method achieves this by using about 20% of the synapses used by other spike algorithms. We also present results of applying our algorithm to classify the MNIST-DVS dataset collected from a real spike-based image sensor and show results comparable to the best reported ones (88.1% accuracy). For VLSI implementations, we show that the reduced synaptic memory can save upto 4X area compared to conventional crossbar topologies. Finally, we also present a biologically realistic spike-based version for calculating the correlations required by the structural learning rule and demonstrate the correspondence between the rate-based and spike-based methods of learning.

No MeSH data available.