第一个尖峰编码促进准确和高效的尖峰神经网络,用于具有丰富时间结构的离散事件
Siying Liu1, Vincent C H Leung1, Pier Luigi Dragotti1
1Communications and Signal Processing Group, Department of Electrical and Electronic Engineering, Imperial College London, London, United Kingdom.
Frontiers in neuroscience
|October 18, 2023
概括
在尖端神经网络 (SNN) 中的第一尖端 (FS) 编码提供了与速率编码 (FR) 相比的能源效率,同时有效地利用尖端时间来处理复杂事件数据. 较长的第一个峰值延迟与更高的分类准确性相关.
科学领域:
- 计算神经科学是一种神经科学.
- 人工智能的人工智能
- 机器学习 机器学习
背景情况:
- 尖端神经网络 (SNN) 擅长处理基于事件的数据.
- 传统的SNN经常使用速率编码 (FR),忽视了精确的峰值时间.
- 时间编码,像时间到第一个峰值 (TTFS),是高效的,但具有挑战性的训练,特别是对现实世界的事件数据,由于不切实际的约束.
研究的目的:
- 为SNN引入一种新的第一尖 (FS) 编码策略,用于分类现实世界的事件序列.
- 在复杂的时间数据中调查第一个峰值时间的意义.
- 在SNN中开发FS编码的强有力的训练方法.
主要方法:
- 为离散的尖峰列车提出了一种新的替代梯度学习方法,以实现FS编码.
- 实现了向前传输,将离散的尖峰时间编码为FS时间.
- 开发了一种使用高斯窗口进行反向传播的错误分配方法,并对尖列车进行监督学习.
主要成果:
- FS编码实现了与FR编码相当的准确性.
- FS编码显示出优越的能源效率.
- FS编码揭示了不同的神经元动态,特别是在具有丰富时间结构的数据上.
- 第一次升的时间延迟较长与分类准确度的提高有关.
结论:
- 第一尖 (FS) 编码是SNN处理复杂时间事件数据的速率编码 (FR) 的可行和高效替代方案.
- 第一个尖峰的时间包含了重要的信息,对于准确的分类至关重要.
- 开发的替代梯度学习方法有效地训练了SNN中的FS编码.
相关概念视频
Integration of Synaptic Events
1.5K
Synaptic integration mainly includes the summation of graded potentials. Graded potentials, regardless of their type, cause subtle alterations in membrane voltage, resulting in either depolarization or hyperpolarization. These incremental changes, when combined or summed, can propel the neuron toward its threshold. Consider, for example, a membrane experiencing a +15 mV shift, causing it to depolarize from -70 mV to -55 mV. In this scenario, graded potentials govern the membrane's ability to...
1.5K
Neural Circuits
1.3K
Neural circuits and neuronal pools are two of the main structures found in the nervous system. Neural circuits are networks of neurons that work together to carry out a specific task or process. They consist of interconnected neurons and glial cells, which provide structural and metabolic support.
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
Neuronal pools are collections of nerve cells with similar functions and interact through chemical and electrical signals. These pools include both interneurons (the central neural circuit nodes that...
1.3K
Propagation of Action Potentials
5.9K
The propagation of an action potential refers to the process by which a nerve impulse, or "action potential," travels along a neuron.
Neurons (nerve cells) have a resting membrane potential, with a slightly negative charge inside compared to outside. This is maintained by ion channels, such as sodium (Na+) and potassium (K+) channels, which control the flow of ions. When a stimulus, like a touch or a signal from another neuron, triggers the neuron, sodium channels open, allowing sodium ions to...
Neurons (nerve cells) have a resting membrane potential, with a slightly negative charge inside compared to outside. This is maintained by ion channels, such as sodium (Na+) and potassium (K+) channels, which control the flow of ions. When a stimulus, like a touch or a signal from another neuron, triggers the neuron, sodium channels open, allowing sodium ions to...
5.9K
Spinal Cord: Information Processing
1.3K
The spinal cord is an integral hub for motor and sensory information that enables the brain to communicate with the peripheral nervous system (PNS). This communication consists of relaying sensory data and transmission of motor commands.
Sensory Information Processing
Sensory information processing begins at the sensory receptors located in the skin and other tissues, which detect somatic sensory stimuli such as touch, temperature, or pain. These receptors function as catalysts, initiating...
Sensory Information Processing
Sensory information processing begins at the sensory receptors located in the skin and other tissues, which detect somatic sensory stimuli such as touch, temperature, or pain. These receptors function as catalysts, initiating...
1.3K


