דגמי מרחב לטנטי מובנים על פצ'ים זמניים רב-סקאליים

תקציר מקורי באנגליתarXiv:2607.19404v1 Announce Type: new Abstract: Multivariate time series encode structural patterns that unfold across multiple temporal scales, yet most forecasting backbones treat learned representations as transient byproducts of prediction, leaving the organizational geometry of these patterns underexploited. We introduce M2Patch, a CNN-based forecasting architecture that maps channel-independent multivariate observations into a structured latent space through two complementary differentiable constraints. Multi-scale patching decomposes the input into overlapping temporal granularities; depthwise separable convolutions with progressive dilation extract scale-specific features in linear time; and per-scale learned projections compress these features into a compact latent representation.