dc.title: Dimensionality reduction of plantar pressure data using principal component analysis: optimal sensor configuration across 12 activities dc.description.abstract: Plantar pressure insoles generate unpractical and hard to manage dimensionality dataset, a single record at 65Hz produces over one million data points per 45 seconds recording session, constraining widespread clinical and athletic deployment. This thesis applies Principal Component Analysis (PCA) to identify the minimum number of sensors required for accurate plantar pressure reconstruction and their optimal anatomical locations across 12 daily and sport-specific activities. 17 healthy participants (7 males, 10 females; age 18–30) performed activities spanning quiet standing, walking, running, stair negotiation, and complex athletic movement patterns. Sensor selection was based on PC1 loading magnitudes explaining 85% of variance accounted for. Gender and task comparison were the primary analytical aim, investigating sex-specific differences in optimal sensor configurations, minimum sensor counts, and pressure distribution patterns across tasks. Findings provide both, a quantitative foundation for designing lightweight, activity-comprehensive wearable insoles for both male and female users and a structured pipeline to identify the most important data depending on task and subject specific loading pattern.