Optimization of Parameter Values for Machine-Learned Models

Abstract

The present disclosure provides computing systems and associated methods for optimizing one or more adjustable parameters (e.g. operating parameters) of a system. In particular, the present disclosure provides a parameter optimization system that can perform one or more black-box optimization techniques to iteratively suggest new sets of parameter values for evaluation. The iterative suggestion and evaluation process can serve to optimize or otherwise improve the overall performance of the system, as evaluated by an objective function that evaluates one or more metrics. The present disclosure also provides a novel black-box optimization technique known as “Gradientless Descent” that is more clever and faster than random search yet retains most of random search's favorable qualities.

1 .- 22 . (canceled) 
     
     
         23 . A computer-implemented method for use in optimization of parameter values for machine-learning models, the method comprising:
 receiving, by one or more computing devices, one or more prior evaluations of performance of a machine learning model, the one or more prior evaluations being respectively associated with one or more prior variants of the machine-learning model, the one or more prior variants of the machine-learning model each having been configured using a different set of adjustable parameter values;   utilizing, by the one or more computing devices, an optimization algorithm to generate a suggested variant of the machine-learning model based at least in part on the one or more prior evaluations of performance and the associated set of adjustable parameter values, the suggested variant of the machine-learning model being defined by a suggested set of adjustable parameter values; and   performing, by the one or more computing devices, transfer learning to obtain initial values for one or more adjustable parameters of the machine-learning model based on the one or more prior variants of the machine-learning model.   
     
     
         24 . The computer-implemented method of  claim 23 , wherein the one or more prior variants of the machine-learning model comprise a plurality of previously optimized machine learned models. 
     
     
         25 . The computer-implemented method of  claim 24 , wherein performing, by the one or more computing devices, transfer learning comprises:
 identifying, by the one or more computing devices, the plurality of previously optimized machine learned models, wherein the plurality of previously optimized machine learned models are organized in a sequence; and   building, by the one or more computing devices, a plurality of Gaussian Process regressors respectively for the plurality of previously optimized machine learned models.   
     
     
         26 . The computer-implemented method of  claim 25 , wherein the Gaussian Process regressor for each previously optimized machine learned model is trained on one or more residuals relative to the Gaussian Process regressor for the previous previously optimized machine learned model in the sequence. 
     
     
         27 . The computer-implemented method of  claim 25 , wherein the sequence is in temporal order based on when the plurality of previously optimized machine learned models were performed. 
     
     
         28 . A computer system operable to suggest parameter values for machine-learned models, the computer system comprising:
 a database that stores one or more results respectively associated with one or more sets of parameter values for one or more adjustable parameters of a machine-learned model, the result for each set of parameter values comprising an evaluation of the machine-learned model constructed with such set of parameter values for the one or more adjustable parameters; one or more processors; and   one or more non-transitory computer-readable media that store instructions that, when executed by the one or more processors, cause the computer system to perform operations, the operations comprising:
 performing one or more black box optimization techniques to generate a suggested set of parameter values for the one or more adjustable parameters of the machine-learned model based at least in part on the one or more results and the one or more sets of parameter values respectively associated with the one or more results; and 
 performing transfer learning to obtain initial parameter values for the one or more adjustable parameters. 
   
     
     
         29 . The computer system of  claim 28 , wherein the operations further comprise:
 accepting an adjustment to the suggested set of parameter values from a user, the adjustment comprising at least one change to the suggested set of parameter values to form an adjusted set of parameter values;   receiving a new result obtained through evaluation of the machine-learned model constructed with the adjusted set of parameter values; and   associating the new result and the adjusted set of parameter values with the one or more results and the one or more sets of parameter values in the database.   
     
     
         30 . The computer system of  claim 29 , wherein the operations further comprise:
 generating a second suggested set of parameter values for the one or more adjustable parameters of the machine-learned model based at least in part on the new result for the adjusted set of parameter values.   
     
     
         31 . The computer system of  claim 28 , wherein performing transfer learning comprises:
 identifying a plurality of previously studied machine-learned models, the plurality of previously studied machine-learned models organized in a sequence; and   building a plurality of Gaussian Process regressors respectively for the plurality of previously studied machine-learned models, wherein the Gaussian Process regressor for each previously studied machine-learned model is trained on one or more residuals relative to the Gaussian Process regressor for a previous previously studied machine-learned model in the sequence.   
     
     
         32 . The computer system of  claim 31 , wherein the sequence is in temporal order based on when the plurality of previously studied machine-learned models were performed. 
     
     
         33 . The computer system of  claim 28 , wherein the one or more adjustable parameters of the machine-learned model comprises one or more adjustable hyperparameters of the machine-learned model. 
     
     
         34 . The computer system of  claim 28 , wherein the operations further comprise performing a plurality of rounds of generation of suggested sets of parameter values using at least two different black box optimization techniques. 
     
     
         35 . The computer system of  claim 34 , wherein the operations further comprise automatically changing black box optimization techniques between at least two of the plurality of rounds of generation of suggested sets of parameter values. 
     
     
         36 . The computer system of  claim 34 , wherein the at least two different black box optimization techniques are stateless so as to enable switching between black box optimization techniques between at least two of the plurality of rounds of generation of suggested sets of parameter values. 
     
     
         37 . The computer system of  claim 28 , wherein the operations further comprise:
 performing a plurality of rounds of generation of suggested sets of parameter values; and   receiving a change to a feasible set of values for at least one of the one or more adjustable parameters of the machine-learned model between at least two of the plurality of rounds of generation of suggested sets of parameter values.   
     
     
         38 . The computer system of  claim 28 , wherein the operations further comprise providing for display a parallel coordinates visualization of the one or more results and the one or more sets of parameter values for the one or more adjustable parameters. 
     
     
         39 . A computer-implemented method to suggest parameter values for machine-learned models, the method comprising:
 receiving, by the one or more computing devices, one or more results respectively associated with one or more sets of parameter values for one or more adjustable parameters of a machine-learned model, the result for each set of parameter values comprising an evaluation of the machine-learned model constructed with such set of parameter values for the one or more adjustable parameters;   generating, by the one or more computing devices, a suggested set of parameter values for the one or more adjustable parameters of the machine-learned model based at least in part on the one or more results and the one or more sets of parameter values respectively associated with the one or more results; and   performing transfer learning to obtain initial parameter values for the one or more adjustable parameters.   
     
     
         40 . The computer-implemented method of  claim 39 , further comprising:
 receiving, by the one or more computing devices, an adjustment to the suggested set of parameter values from a user, the adjustment comprising at least one change to the suggested set of parameter values to form an adjusted set of parameter values;   receiving, by the one or more computing devices, a new result associated with the adjusted set of parameter values; and   associating, by the one or more computing devices, the new result and the adjusted set of parameter values with the one or more results and the one or more sets of parameter values.   
     
     
         41 . The computer-implemented method of  claim 40 , further comprising:
 generating, by the one or more computing devices, a second suggested set of parameter values for the one or more adjustable parameters of the machine-learned model based at least in part on the new result for the adjusted set of parameter values.   
     
     
         42 . The computer-implemented method of  claim 39 , wherein the one or more adjustable parameters of the machine-learned model comprises one or more adjustable hyperparameters of the machine-learned model.