Synthetic_Multimodal_Dataset_Dyslexia.csv

This dataset is <b>synthetic</b>, generated for the purpose of our on-going research titled <i>"Dyslexia Detection Using Multi-Modality Data."</i> The dataset was created to simulate realistic scenarios by incorporating key characteristics of individuals with dyslexia (abnormal group) and those without dyslexia (normal group).The generation process was guided by:<b>Known Characteristics of Dyslexic Individuals:</b>Handwriting patterns, including irregular pressure, inconsistent stroke velocity, and wider variability in letter spacing.Distinct EEG activity patterns, such as heightened theta and altered alpha frequencies.Behavioral indicators like slower eye blink rates and reduced smile scores during cognitive tasks.<b>Control Characteristics for Normal Individuals:</b>More consistent handwriting patterns and EEG signal distributions.Behavioral indicators aligned with typical age-appropriate cognitive and emotional responses.This synthetic dataset, therefore, allows us to explore multimodal patterns relevant to dyslexia detection while ensuring complete privacy and ethical compliance.Research UtilizationThe dataset forms the foundation of our study, enabling the development of machine learning models that can:Differentiate between dyslexic and non-dyslexic individuals.Leverage multimodal features for improved accuracy and robustness in dyslexia detection.本数据集为合成数据集（synthetic），旨在支撑我们正在开展的题为《基于多模态数据的阅读障碍检测》（"Dyslexia Detection Using Multi-Modality Data"）的研究工作。本数据集通过纳入阅读障碍人群（异常组）与非阅读障碍人群（正常组）的核心特征，以模拟真实应用场景。 数据集的生成遵循以下依据： ### 阅读障碍人群的已知特征： 书写模式异常，如笔压不均、笔画速度不稳定、字母间距波动幅度更大；独特的脑电（EEG）活动模式，如θ波活动增强、α波频率异常；行为指标异常，如认知任务中眨眼速率偏低、微笑评分较低。 ### 正常人群的对照特征： 书写模式与脑电信号分布更为规整一致；行为指标符合对应年龄段的典型认知与情绪反应模式。 因此，该合成数据集可支持我们探索与阅读障碍检测相关的多模态特征模式，同时可充分保障数据隐私与伦理合规性。 ## 研究应用 本数据集作为本研究的核心基础，可用于开发能够实现以下功能的机器学习模型： 1. 区分阅读障碍人群与非阅读障碍人群； 2. 利用多模态特征提升阅读障碍检测的准确率与鲁棒性。