Comparing Teaching Strategies of a Machine Learning-based Prosthetic Arm

The dataset contains EMG data collected with a Myo Armband during an experiment comparing strategies to teach a machine to classify muscle contraction patterns (gestures). The data files are structured as follows: there are three folders, one for each condition in the experiment, namely Learner-led condition(LLC), Teacher-led condition (TLC) ,Random condition (RC). Each folder contains folders for 17 participants, denoted as Px , where x is the participant ID. Each participant's folder contains three database files, the training data (instances-training-x), the post-test data (instances-posttest-x), and the positive-negative examples of phase 4(instances-pos-neg-x).

本数据集包含使用Myo臂带（Myo Armband）采集的肌电（Electromyography, EMG）数据，其采集自一项对比不同教学策略的实验，该实验旨在实现机器对肌肉收缩模式（即手势）的分类。数据文件的组织结构如下：共设置三个文件夹，分别对应实验中的三类实验条件，即学习者主导条件（Learner-led condition, LLC）、教师主导条件（Teacher-led condition, TLC）与随机条件（Random condition, RC）。每个条件文件夹下均包含17个参与者的子文件夹，命名格式为Px，其中x为参与者编号。每个参与者的子文件夹内存储有三个数据库文件：训练数据集（instances-training-x）、后测数据集（instances-posttest-x）以及第四阶段的正负样本数据集（instances-pos-neg-x）。