On the Performance of Machine Learning Methods for Breakthrough Curve Prediction

Goal‐Oriented Two‐Layered Kernel Models as Automated Surrogates for Surface Kinetics in Reactor Simulations

“…The in-situ adaptive tabulation [13,19,21,22,[24][25][26] builds such a map during the simulation by exploiting prior solutions to estimate new queries. In contrast, surrogate models like splines or the (error-based modified) Shepard interpolation approach map precomputed solutions to accelerate reactor simulations [2,15,16,18,20,[27][28][29][30][31] or even spatial subsystems of the reactor [32,33] and breakthrough curves [34]. Lately, primarily machine learning techniques like random forests [35,36] or neural networks [2,3,37] have been used for accurate predictions of steady-state surface kinetics because they can overcome the so-called curse of dimensionality [38], i.e., the exponentially increasing difficulty to learn high-dimensional data.…”

“…While other methods estimate their prediction error by the distance to known points [25] or the slope of the func- and breakthrough curves [34]. Lately, primarily machine learning techniques like random forests [35,36] or neural networks [2,3,37] have been used for accurate predictions of steady-state surface kinetics because they can overcome the so-called curse of dimensionality [38], i.e., the exponentially increasing difficulty to learn high-dimensional data.…”

“…Here, the common kernel layer is extended by an additional linear layer, which adjusts the kernel to the magnitude and importance of the features (here: reaction conditions) and thereby significantly increases the model accuracy. The parameters of this linear transformation are obtained by gradient descent optimization of small kernel models on and breakthrough curves [34]. Lately, primarily machine learning techniques like random forests [35,36] or neural networks [2,3,37] have been used for accurate predictions of steady-state surface kinetics because they can overcome the so-called curse of dimensionality [38], i.e., the exponentially increasing difficulty to learn high-dimensional data.…”