Development of elemental models for the brain simulator

我们一直致力于神经元输入输出关系的建模。我们发表了一篇论文，描述了我们的MAT（多时间尺度自适应阈值）模型，我们在预测真实的生物神经元的尖峰时间的国际比赛中获得了第一名（小林，Tsubo和Shinomoto，2009）。我们进一步成功地增强了MAT模型，使其可以代表Izhikevich所代表的非平稳输入的神经元放电模式的可能变化，同时保持了原始MAT模型固有的高预测能力（Yamauchi，Kim和Shinomoto，2011）。此外，我们邀请了杰出的研究人员进行讨论和讲座，并开始了国际合作研究。

We have been working for modeling neuronal input-output relationship. We published a paper describing our MAT (multi-timescale adaptive threshold) model with which we won the first places in international competitions for predicting spike times of real biological neurons (Kobayashi, Tsubo, and Shinomoto, 2009). We further succeeded in augmenting the MAT model, so that it may represent the possible variety of neuronal firing patterns for nonstationary input, as represented by Izhikevich, while keeping the high predictive ability inherent in the original MAT model (Yamauchi, Kim, and Shinomoto, 2011). In addition, we invited distinguished researchers for discussion and lectures, and started international collaboration studies.