A review on data clustering using spiking neural network (SNN) models

Mohamed, Siti Aisyah; Othman, Muhaini; Afifi, Mohd Hafizul

doi:10.11591/ijeecs.v15.i3.pp1392-1400

Cited by 10 publications

(5 citation statements)

References 19 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More importantly, the relatively low migration energy barrier of V Sn results in their easy mobility under electrical stimuli, bringing about dynamical correlation between V Sn and Ag, thus increasing the variation of the threshold switching voltage. − By exploiting the intrinsic stochasticity in the TS device, a stochastic LIF neuron is demonstrated by integrating the TS memory with a simple RC circuit. , The measured value of the firing threshold shows variability, which is extracted and implemented to add the dropout function in an SNN for MNIST handwritten digit recognition with reduced overfitting problems. Additionally, by applying longer voltage pulses to enhance the temporal correlation of the current responses in the device, nociceptive behaviors of “threshold”, “relaxation”, no adaptation” and “sensitization” are successfully emulated. − These results suggest that the ionic conductive 2D SnSe holds significant potential in the application of future multifunctional neuromorphic devices. − …”

Section: Introductionmentioning

confidence: 93%

Threshold Switching Memristor Based on 2D SnSe for Nociceptive and Leaky-Integrate and Fire Neuron Simulation

Qin,

Wu,

et al. 2024

ACS Appl. Electron. Mater.

View full text Add to dashboard Cite

Multifunctional neuromorphic devices to tackle complex tasks are highly desirable for the development of artificial neural networks. Threshold switching (TS) memory, which exhibits volatile abrupt resistance change under external electric fields, is capable of emulating multiple biological behaviors because of its rich temporal dynamics. Here, a TS device based on two-dimensional (2D) SnSe is demonstrated. Owing to the diffusive dynamics of Ag ions in SnSe, intrinsic stochasticity of the TS behavior is observed, which can be exploited to construct a compact stochastic Leaky-Integrate and Fire (LIF) model with improved performance in spiking neuron network (SNN). Moreover, an artificial nociceptor is constructed based on the 2D TS device, successfully emulating typical nociceptive features of "threshold", "relaxation", "no adaptation", "hyperalgesia" and "allodynia". The realization of bioinspired devices with combined sensory and information processing abilities paves the way for developing neuromorphic electronics for SNN and humanoid robots.

show abstract

Section: Introductionmentioning

confidence: 93%

Threshold Switching Memristor Based on 2D SnSe for Nociceptive and Leaky-Integrate and Fire Neuron Simulation

Qin,

Wu,

et al. 2024

ACS Appl. Electron. Mater.

View full text Add to dashboard Cite

show abstract

“…Typical SNN models include Hodgkin-Huxley (HH) model, Integrate-and-fire (IF) model, Izhikevich (IZH) model, the leak-integrate-and-fire (LIF) model, etc [5]. LIF is the most widely used and relatively simple neuron model.…”

Section: Preliminaries Snnmentioning

confidence: 99%

Adaptive Drop Approaches to Train Spiking-YOLO Network for Traffic Flow Counting

Zhang,

Wu,

Liu

et al. 2024

Preprint

View full text Add to dashboard Cite

Traffic flow counting is an object detection problem. YOLO (" You Only Look Once ") is a popular object detection network. Spiking-YOLO converts the YOLO network from ANN(artificial neural network) to SNN (spiking neural network), which consumes less power than YOLO. However, Spiking-YOLO is difficult to train and has poor accuracy because of overfitting. Dropout and DropConnect are two methods to regularize neural network models and have produced cutting-edge outcomes in many benchmarks. The two drop approaches are first applied to Spiking-YOLO. Considering that a higher membrane potential of a biological neuron implies a higher probability of neural activation, we propose three adaptive drop algorithms—Spiking-YOLO with adaptive Dropout, Spiking-YOLO with adaptive DropConnect, and Spiking-YOLO with group adaptive Drop, which adaptively adjusts the keep probability for training SNNs. The experimental results show that the adaptive algorithms have fast convergence speed with high detection accuracy and mAP value, and work well in the case of high-speed streaming video, making them suitable for online learning scenarios.

show abstract

“…TF-IDF is one that only multiplies term frequency and inverse document frequency that TF is the number of appearances of a term in a document and IDF reduces the dominant terms that often appear in various documents or files, by calculating the inverse frequency of them that contain a word and can exclude a collection of words [1]- [3]. TF-IDF is commonly used in the classification process in existing documents and was studied by several classification methods [4]- [6].…”

Section: Introductionmentioning

confidence: 99%

The effect of the TF-IDF algorithm in times series in forecasting word on social media

Lubis

Nasution

Sitompul

et al. 2021

IJEECS

View full text Add to dashboard Cite

Forecasting is one of the main topics in data mining or machine learning in which forecasting, a group of data used, has a label class or target. Thus, many algorithms for solving forecasting problems are categorized as supervised learning with the aim of conducting training. In this case, the things that were supervised were the label or target data playing a role as a 'supervisor' who supervise the training process in achieving a certain level of accuracy or precision. Time series is a method that is generally used to forecast based on time and can forecast words in social media. In this study had conducted the word forecasting on twitter with 1734 tweets which were interpreted as weighted documents using the TF-IDF algorithm with a frequency that often comes out in tweets so the TF-IDF value is getting smaller and vice versa. After getting the word weight value of the tweets, a time series forecast was performed with the test data of 1734 tweets that the results referred to 1203 categories of Slack words and 531 verb tweets as training data resulting in good accuracy. The division of word forecasting was classified into two groups i.e. inactive users and active users. The results obtained were processed with a MAPE calculation process of 50% for inactive users and 0.1980198% for active users.

show abstract

A review on data clustering using spiking neural network (SNN) models

Cited by 10 publications

References 19 publications

Threshold Switching Memristor Based on 2D SnSe for Nociceptive and Leaky-Integrate and Fire Neuron Simulation

Threshold Switching Memristor Based on 2D SnSe for Nociceptive and Leaky-Integrate and Fire Neuron Simulation

Adaptive Drop Approaches to Train Spiking-YOLO Network for Traffic Flow Counting

The effect of the TF-IDF algorithm in times series in forecasting word on social media

Contact Info

Product

Resources

About