Free Access

Fig. 5.


The fits for time series using LASSO and random forests for Rm = 1.5 × 104. Just like the fits for spatial data in Fig. 3, we find that LASSO does considerably better than random forests, the latter of which again has a step like shape. Left: we show the time series only up to the kinematic phase (∼0.01 resistive times) with the inset describing the train/test split. Right: time series is between kinematic and saturation phases. LASSO seems to capture the shape of the curve but is offset. In both cases, random forest prediction returns a nearly horizontal line, characteristic of decision tree regression. The black vertical dashed line presents the train-test split (80% training data).

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.