datafiles for incremental research.rar

Wall following robot navigation data. <br> This file contain all the data used, for sensor 2, 4, 24 file, the original files are available in: Freire, Veloso, &amp; Barreto. (2009). UCI Machine Learning Repository: Wall-Following Robot Navigation Data Data Set. UCI Machine Learning Repository: Wall-Following Robot Navigation Data Data Set. Retrieved 2017, from https://archive.ics.uci.edu/ml/datasets/Wall-Following+Robot+Navigation+Data in our research we reduced all original files using Condensed Nearest Neighbors CNN instance selection algorithm We used reduced files as training files, we removed all training instances from the original file and used the remaining data as test files to avoid data leakage. <br> for gathered data, we used the same method of gathering as the other files, labelled them, reduced them using CNN then finally prepared separate training and test files. <br> for additional information please contact: [e-mail]

循壁机器人导航数据集。<br>本文件包含本次研究所用的全部数据。针对传感器2、4、24对应的文件，其原始数据源自Freire、Veloso与Barreto（2009年）的研究成果：UCI机器学习库（UCI Machine Learning Repository）中的《循壁机器人导航数据集》，2017年检索自https://archive.ics.uci.edu/ml/datasets/Wall-Following+Robot+Navigation+Data。<br>本研究中，我们采用压缩最近邻（Condensed Nearest Neighbors, CNN）实例选择算法对所有原始文件进行了约简处理，将约简后的文件用作训练集；同时为避免数据泄露，我们从原始文件中移除全部训练样本，将剩余数据作为测试集。<br>对于本次采集的数据集，我们采用与其他文件一致的采集流程，完成标注后使用CNN算法进行约简，最终构建出独立的训练集与测试集。<br>如需获取更多信息，请联系邮箱：[e-mail]