[1909.11822v1] DisCo: Physics-Based Unsupervised Discovery of Coherent Structures in Spatiotemporal Systems
We also demonstrate state-of-the-art segmentation results for three complex scientific datasets, including on 89.5 TB of simulated climate data (lightcones) in 6.6 minutes endto-end

Abstract: Extracting actionable insight from complex unlabeled scientific data is an
open challenge and key to unlocking data-driven discovery in science.
Complementary and alternative to supervised machine learning approaches,
unsupervised physics-based methods based on behavior-driven theories hold great
promise. Due to computational limitations, practical application on real-world
domain science problems has lagged far behind theoretical development. We
present our first step towards bridging this divide - DisCo - a
high-performance distributed workflow for the behavior-driven local causal
state theory. DisCo provides a scalable unsupervised physics-based
representation learning method that decomposes spatiotemporal systems into
their structurally relevant components, which are captured by the latent local
causal state variables. Complex spatiotemporal systems are generally highly
structured and organize around a lower-dimensional skeleton of coherent
structures, and in several firsts we demonstrate the efficacy of DisCo in
capturing such structures from observational and simulated scientific data. To
the best of our knowledge, DisCo is also the first application software
developed entirely in Python to scale to over 1000 machine nodes, providing
good performance along with ensuring domain scientists' productivity. We
developed scalable, performant methods optimized for Intel many-core processors
that will be upstreamed to open-source Python library packages. Our capstone
experiment, using newly developed DisCo workflow and libraries, performs
unsupervised spacetime segmentation analysis of CAM5.1 climate simulation data,
processing an unprecedented 89.5 TB in 6.6 minutes end-to-end using 1024 Intel
Haswell nodes on the Cori supercomputer obtaining 91% weak-scaling and 64%
strong-scaling efficiency.

‹Fig. 1. 2+1D lightcone template with past horizon h− = 2, future horizon h+ = 2, speed of information propagation c = 1. (Description of the DisCo project)Fig. 2. Breakdown of execution time spent in various stages of the DisCo on Haswell nodes with K-Means. Left : weak scaling and Right: strong scaling. Parallel efficiency are plotted on the secondary axis. (Multi-node scaling)

ell__plus * r, t == (X * r __deriv, t__deriv : t < t__deriv <= t__plus, / r * __deriv - r / <= c(t__deriv - t)) ~

Fig. 3. Breakdown of execution time spent in various stages of the DisCo on Haswell and KNL nodes with DBSCAN. Left : weak scaling and Right: strong scaling. Parallel efficiency are plotted on the secondary axis. (Single-node performance)Fig. 4. Structural segmentation results for the three scientific data sets using K-Means lightcone clustering. The leftmost image of each row shows a snapshot from the data spacetime fields, and the other image(s) in the row show corresponding snapshots from the reconstructed local causal state spacetime fields. Reconstruction parameters given as (h− , h+ , c, K− , τ): (b) - (14, 2, 1, 10, 0.8), (c) - (14, 2, 1, 4, 0.0), (e) - (3, 3, 3, 8, 0), (g) - (3, 3, 1, 16, 0.04). K+ = 10 and 0.05 for chi-squared significance level were used for all reconstructions. Full segmentation videos are available on the DisCo YouTube channel [71] (Hero Run)Fig. 5. Comparison of structural segmentation results on 2D turbulence using DBSCAN (a) and K-Means (b) for lightcone clustering. The K-Means results in (b) are the same as Figure ?? (b), repeated here for easier comparison. The DBSCAN results shown in (a) use reconstruction parameters (h− , h+ , c) = (3, 2, 1), τ = 0.0, eps = 0.0, and minpts = 10. (Reconstruction Parameters)›