首页> 外文期刊>International journal of mathematics, game theory and algebra >APPLICATION OF HIDDEN MARKOV MODELS FOR ESTIMATING THE ESTRUS AND NONESTRUS BEHAVIORS OF SHIBA GOATS
【24h】

APPLICATION OF HIDDEN MARKOV MODELS FOR ESTIMATING THE ESTRUS AND NONESTRUS BEHAVIORS OF SHIBA GOATS

机译:隐马尔可夫模型在山羊估算中的应用和非estrus行为

获取原文
获取原文并翻译 | 示例
           

摘要

Hidden Markov models (HMMs) have been used in various fields such as speech recognition and the estimation of human behaviors, but they have not been applied to animal husbandry in previous studies. Mammalian behavior has typically been studied based on observations, but direct observation requires substantial time and effort if the number of mammals observed is high or the observations are obtained over a long period of time. Thus, a machine-learning method such as an HMM may be suitable for the evaluation of mammalian behavior. Therefore, this study determined the feasibility of using an HMM for monitoring the behavior of goats as a typical mammalian subject, where it was used to assess whether goats were in estrus. In total, the model was used to test whether 16 goats were in estrus (n = 8) or not in estrus (n = 8). Tracking data were obtained with a video tracking system and used to determine whether the goats were in various states, i.e., "nonestrus", or "estrus". According to the percentage concordance defined as an index of the match between human observations and the estimates obtained by the HMM, the results showed that the HMM successfully simulated the human observations.
机译:隐藏的马尔可夫模型(HMMS)已被用于语音识别和人类行为的估计等各种领域,但他们尚未应用于以前研究的畜牧业。通常基于观察结果研究哺乳动物行为,但如果观察到的哺乳动物的数量或在很长一段时间内获得观察结果,则直接观察需要大量的时间和努力。因此,诸如HMM的机器学习方法可以适合于评估哺乳动物行为。因此,这项研究确定了使用HMM监测山羊作为典型哺乳动物主题的行为的可行性,用于评估山羊是否在雌性中。总共可以使用该模型来测试16只山羊是否在Estrus(n = 8)中(n = 8)。使用视频跟踪系统获得跟踪数据,并用于确定山羊是否处于各种状态,即“非estrus”或“Estrus”。根据被定义为人类观察与嗯,通过肝脏获得的估计的匹配指标的百分比,结果表明,嗯成功模拟人类观察。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号