New Classification of Collective Animal Behaviour as an Autonomous System
New Classification of Collective Animal Behaviour as an Autonomous System
复制标题
作为自治系统的集体动物行为的新分类
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
Toshiki Fukushima
中科院分区：
文献类型：
--
作者：
T. Niizato;Kotaro Sakamoto;Yoh;H. Murakami;Takenori Tomaru;Tomotaro Hoshika;Toshiki Fukushima
综合信息理论(IIT)最初被用来描述人类意识的内在-因果大脑网络结构。这一理论可能被用来概念化复杂的生命系统。在之前的一项研究中，我们分析了高山舌鱼的集体行为。我们发现，IIT 3.0在$phi$值(即群体完整性)方面，在三到四个鱼群之间表现出质的不连续性。其他指标，如相互信息，则没有表现出这样的特征。在这项研究中，我们继续我们之前的发现，并引入了两个新的因素。首先，我们定义全局参数设置以确定不同类型的组完整性。其次，我们设置了几个时间尺度(从$\Delta t=5/120$S到$\Delta t=120/120$S)。结果表明，尽管群体规模很小，但我们成功地根据群体完整性程度对鱼群进行了分类。具体的分类包括两鱼鱼群的追随者，三鱼鱼群的裂变-融合，四鱼鱼群的领导层的出现，以及五鱼鱼群的类Boid行为的出现。这些细微的分类以前从未被揭示过。最后，我们讨论了集体行为中一个长期存在的悖论，即所谓的堆悖论，通过我们的IIT分析，可以为这个悖论提供两个试探性的答案。
Integrated information theory (IIT) was initially proposed to describe human consciousness in terms of intrinsic-causal brain network structures. This theory could potentially be used for conceptualising complex living systems. In a previous study, we analysed collective behaviour in {\it Plecoglossus altivelis}. We found that IIT 3.0 exhibits qualitative discontinuity between three and four schools of fish in terms of $\Phi$ values (i.e., group integrity). Other measures, such as mutual information, did not show such characteristics. In this study, we follow up on our previous findings and introduce two new factors. First, we define the global parameter settings to determine a different kind of group integrity. Second, we set several timescales (from $\Delta t =5/120$ s to $\Delta t =120/120$ s). The results showed that we succeeded in classifying fish school according to their group size in terms of the degree of group integrity, despite the small group size. The concrete classification includes the followership for a two-fish school, fission--fusion for a three-fish school, emergence of leadership for a four-fish school, and emergence of Boid-like behaviour for a five-fish school. These minute classifications have never been revealed before. Finally, we discuss one of the longstanding paradoxes in collective behaviour, known as the heap paradox, for which two tentative answers could be provided through our IIT analysis.