Robustness of Multidimensional Optimization Outcomes: A General Approach and a Case Study

In multidimensional parameter optimization of complex systems, the preferred solution must also be robust to virtually inevitable perturbations and uncertainties. Having a conceptually simple and computationally facile metric that can help distinguish between candidate optimum solutions in a post processing step is useful. Motivated by free energy function in statistical physics, which evaluates a trade-off between entropy and enthalpy, here we introduce a novel statistical robustness metric that assesses robustness with respect to possible to inevitable uncertainties in the objective function values or optimal parameters. The metric is the expected value ⟨E⟩ of the objective function, evaluated using weighted samples in a box around each optimum. The width of the sample distribution is problem-specific. As a case study, the proposed robustness metric is employed to find the most robust optimal solution in 5-dimensional parameter space in the context of dielectric boundary optimization in atomistic modeling, relevant to computational drug discovery.

[1]  Jianying Hu,et al.  Artificial Intelligence for Clinical Trial Design. , 2019, Trends in pharmacological sciences.

[2]  Hideaki Shimazaki,et al.  The principles of adaptation in organisms and machines I: machine learning, information theory, and thermodynamics , 2019, 1902.11233.

[3]  Yao Zhang,et al.  Energy–entropy competition and the effectiveness of stochastic gradient descent in machine learning , 2018, Molecular Physics.

[4]  Saeed Izadi,et al.  Water models for biomolecular simulations , 2018 .

[5]  Andrea Castelletti,et al.  Robustness Metrics: How Are They Calculated, When Should They Be Used and Why Do They Give Different Results? , 2018 .

[6]  Robert C. Harris,et al.  Accuracy Comparison of Generalized Born Models in the Calculation of Electrostatic Binding Free Energies. , 2018, Journal of chemical theory and computation.

[7]  A. Onufriev,et al.  Grid-Based Surface Generalized Born Model for Calculation of Electrostatic Binding Free Energies , 2017, J. Chem. Inf. Model..

[8]  Raphael T. Haftka,et al.  Balancing diversity and performance in global optimization , 2016 .

[9]  Jan H. Kwakkel,et al.  How Robust is a Robust Policy? Comparing Alternative Robustness Metrics for Robust Decision-Making , 2016 .

[10]  A. Onufriev,et al.  Protein-Ligand Electrostatic Binding Free Energies from Explicit and Implicit Solvation. , 2015, Journal of chemical theory and computation.

[11]  Ali Mohades,et al.  Structure-Based Analysis of Protein Binding Pockets Using Von Neumann Entropy , 2014, ISBRA.

[12]  Anirban Chaudhuri,et al.  Global Search for Diverse Competitive Designs , 2014 .

[13]  A. Onufriev,et al.  Accuracy of continuum electrostatic calculations based on three common dielectric boundary definitions. , 2014, Journal of theoretical & computational chemistry.

[14]  Boris Aguilar,et al.  Efficient Computation of the Total Solvation Energy of Small Molecules via the R6 Generalized Born Model. , 2012, Journal of chemical theory and computation.

[15]  Alexey V. Onufriev,et al.  Continuum Electrostatics Solvent Modeling with the Generalized Born Model , 2010 .

[16]  L. Watson,et al.  Algorithm 897: VTDIRECT95: Serial and parallel codes for the global optimization algorithm direct , 2009, TOMS.

[17]  Masha Sosonkina,et al.  Performance Modeling and Analysis of a Massively Parallel Direct—Part 1 , 2009, Int. J. High Perform. Comput. Appl..

[18]  Masha Sosonkina,et al.  Performance Modeling and Analysis of a Massively Parallel Direct—Part 2 , 2009, Int. J. High Perform. Comput. Appl..

[19]  M. Gilson,et al.  Calculation of protein-ligand binding affinities. , 2007, Annual review of biophysics and biomolecular structure.

[20]  Layne T. Watson,et al.  Numerical integration in statistical decision-theoretic methods for robust design optimization , 2008 .

[21]  Richard J. Beckman,et al.  A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output From a Computer Code , 2000, Technometrics.

[22]  C. D. Perttunen,et al.  Lipschitzian optimization without the Lipschitz constant , 1993 .

[23]  W. L. Jorgensen,et al.  Comparison of simple potential functions for simulating liquid water , 1983 .