【講題】Self-supervised learning reduces label noise in sharp wave ripple classification

【演講時間】4月17日(星期五)下午1點30分  

【演講地點】清華大學校本部第二綜合大樓B側8樓A813室

【摘要】

In electrophysiological signal analysis, label noise caused by human error or subjectivity frequently compromises the accuracy of time-series classification. This critically impairs the reliability of data classification, presenting significant barriers to extracting meaningful insights. This study addresses the issue by applying self-supervised learning (SSL) to the classification of sharp wave ripples (SWRs), which are high-frequency oscillations critical to memory processing. By utilizing SSL, we effectively relabel SWR data, leveraging the inherent structural patterns within time-series data to improve label quality without relying on external labeling. The application of SSL to SWR datasets has yielded a 10% increase in classification accuracy. The study's findings suggest the transformative capability of SSL in improving data quality across various domains reliant on precise time-series data classification.