Abstract: A method for process control using predictive long short term memory includes obtaining historical post-process measurements taken on prior products of the manufacturing process; obtaining historical in-process measurements taken on prior workpieces during the manufacturing process; training a neural network to predict each of the historical post-process measurements, in response to the corresponding historical in-process measurements and preceding historical post-process measurements; obtaining present in-process measurements on a present workpiece during the manufacturing process; predicting a future post-process measurement for the present workpiece, by providing the present in-process measurements and the historical post-process measurements as inputs to the neural network; and adjusting at least one controllable variable of the manufacturing process in response to the prediction of the future post-process measurement.

Abstract: Techniques for generating model ensembles are provided. A plurality of models trained to generate predictions at each of a plurality of intervals is received. A respective prediction accuracy of each respective model of the plurality of models is determined for a first interval of the plurality of intervals by processing labeled evaluation data using the respective model. Additionally, a model ensemble specifying one or more of the plurality of models for each of the plurality of intervals is generated, comprising selecting, for the first interval, a first model of the plurality of models based on (i) the respective prediction accuracies and (ii) at least one non-error metric.

Abstract: A method of improving at least one of quality and yield of a physical process comprises: obtaining values, from respective performances of the physical process, for a plurality of variables associated with the physical process; determining at least one Gaussian mixture model (GMM) representing the values for the variables for the performances of the physical process; based at least in part on the at least one GMM, computing at least one anomaly score for at least one of the variables for at least one of the performances of the physical process; based on the at least one anomaly score, identifying the at least one of the performances of the physical process as an outlier; and, based at least in part on the outlier identification, modifying the at least one of the variables for one or more subsequent performances of the physical process.

Abstract: In an approach to hyperparameter optimization, one or more computer processors express a hyperparameter tuning process of a model based on a type of model, one or more dimensions of a training dataset, associated loss function of the model, and associated computational constraints of the model, comprising: identifying a set of optimal hyper-rectangles based a calculated local variability and a calculated best function value; calculating a point as a representative for each identified potentially optimal hyper-rectangle by locally searching over the identified set of potentially optimal hyper-rectangles; dividing one or more hyper-rectangles in the identified set of optimal hyper-rectangles into a plurality of smaller hyper-rectangles based on each calculated point; and calculating one or more optimal hyperparameters utilizing a globally converged hyper-rectangle from the plurality of smaller hyper-rectangles.

Abstract: Aspects of the invention include an optimal interpretable decision tree using integer linear programming techniques. A non-limiting example computer-implemented method includes receiving, using a processor, a plurality of data inputs from a process and selecting, using the processor, a data subset from the plurality of data inputs by solving linear programming to obtain a solution. The method builds and optimizes, using the processor, an optimal decision tree based on the data subset and alerts, using the processor, a user when a prediction of the optimal decision tree is greater than a threshold value.

Abstract: Aspects of the invention include implemented method includes selecting an optimization algorithm for the control system of a processing plant based on whether the control system is guided by a linear-based predictive model or a non-linear-based predictive model, in which a gradient is available. Calculating set-point variables using the optimization algorithm. Predicting an output based on the calculated set-point variables. Comparing an actual output at the processing plant to the predicted output. Suspending a physical process at the processing plant in response to the actual output being a threshold value apart from the predicted output.

Abstract: A computer-implemented method for a shuffling-type gradient for training a machine learning model using a stochastic gradient descent (SGD) includes the operations of uniformly randomly distributing data samples or coordinate updates of a training data, and calculating the learning rates for a no-shuffling scheme and a shuffling scheme. A combined operation of the no-shuffling scheme and the shuffling scheme of the training data is performed using a stochastic gradient descent (SGD) algorithm. The combined operation is switched to performing only the shuffling scheme from the no-shuffling scheme based on one or more predetermined criterion; and training the machine learning models with the training data based on the combined no-shuffling scheme and shuffling scheme.

Abstract: Aspects of the invention include training an optimal interpretable decision tree for regression using mixed-integer linear programming techniques. A non-limiting example computer-implemented method includes receiving, using a processor, input data that includes time-series data. The method further includes training, using a binary mixed-integer linear program of the processor, an ODT for regression based on the input data. During the training process one or more outliers are filtered out by a linear loss model that minimizes training loss and outlier loss.

Abstract: Aspects of the invention include implemented method includes selecting an optimization algorithm for the control system of a processing plant based on whether the control system is guided by a linear-based predictive model or a non-linear-based predictive model, in which a gradient is available. Calculating set-point variables using the optimization algorithm. Predicting an output based on the calculated set-point variables. Comparing an actual output at the processing plant to the predicted output. Suspending a physical process at the processing plant in response to the actual output being a threshold value apart from the predicted output.

Abstract: A computer-implemented method for maintaining equipment in a geo-distributed system includes receiving, by a processor, a selection of quantities to optimize when adjusting a maintenance schedule of the geo-distributed system that includes multiple pieces of equipment that are spread over a geographical region, and wherein the maintenance schedule identifies when a set of maintenance tasks are executed at a first equipment from the geo-distributed system over a predetermined duration. The method further includes generating, by the processor, a mixed-integer linear program for optimizing the maintenance schedule using a set of predetermined constraints. The method further includes executing, by the processor, the mixed-integer linear program via a mixed-integer linear program solver. The method further includes adjusting, by the processor, the maintenance schedule by selecting only a subset of the maintenance tasks.

Abstract: In an approach to hyperparameter optimization, one or more computer processors express a hyperparameter tuning process of a model based on a type of model, one or more dimensions of a training dataset, associated loss function of the model, and associated computational constraints of the model, comprising: identifying a set of optimal hyper-rectangles based a calculated local variability and a calculated best function value; calculating a point as a representative for each identified potentially optimal hyper-rectangle by locally searching over the identified set of potentially optimal hyper-rectangles; dividing one or more hyper-rectangles in the identified set of optimal hyper-rectangles into a plurality of smaller hyper-rectangles based on each calculated point; and calculating one or more optimal hyperparameters utilizing a globally converged hyper-rectangle from the plurality of smaller hyper-rectangles.

Abstract: Techniques for model evaluation and selection are provided. A plurality of models trained to generate predictions at each of a plurality of intervals is received, and a plurality of model ensembles, each specifying one or more of the plurality of models for each of the plurality of intervals, is generated. A test data set is received, where the test data set includes values for at least a first interval of the plurality of intervals and does not include values for at least a second interval of the plurality of intervals. A first model ensemble, of the plurality of model ensembles, is selected based on processing the test data set using each of the plurality of model ensembles.

Abstract: Techniques for generating model ensembles are provided. A plurality of models trained to generate predictions at each of a plurality of intervals is received. A respective prediction accuracy of each respective model of the plurality of models is determined for a first interval of the plurality of intervals by processing labeled evaluation data using the respective model. Additionally, a model ensemble specifying one or more of the plurality of models for each of the plurality of intervals is generated, comprising selecting, for the first interval, a first model of the plurality of models based on (i) the respective prediction accuracies and (ii) at least one non-error metric.

Abstract: Aspects of the invention include an optimal interpretable decision tree using integer linear programming techniques. A non-limiting example computer-implemented method includes receiving, using a processor, a plurality of data inputs from a process and selecting, using the processor, a data subset from the plurality of data inputs by solving linear programming to obtain a solution. The method builds and optimizes, using the processor, an optimal decision tree based on the data subset and alerts, using the processor, a user when a prediction of the optimal decision tree is greater than a threshold value.

Abstract: A method of improving at least one of quality and yield of a physical process comprises: obtaining values, from respective performances of the physical process, for a plurality of variables associated with the physical process; determining at least one Gaussian mixture model (GMM) representing the values for the variables for the performances of the physical process; based at least in part on the at least one GMM, computing at least one anomaly score for at least one of the variables for at least one of the performances of the physical process; based on the at least one anomaly score, identifying the at least one of the performances of the physical process as an outlier; and, based at least in part on the outlier identification, modifying the at least one of the variables for one or more subsequent performances of the physical process.

Abstract: A method for process control using predictive long short term memory includes obtaining historical post-process measurements taken on prior products of the manufacturing process; obtaining historical in-process measurements taken on prior workpieces during the manufacturing process; training a neural network to predict each of the historical post-process measurements, in response to the corresponding historical in-process measurements and preceding historical post-process measurements; obtaining present in-process measurements on a present workpiece during the manufacturing process; predicting a future post-process measurement for the present workpiece, by providing the present in-process measurements and the historical post-process measurements as inputs to the neural network; and adjusting at least one controllable variable of the manufacturing process in response to the prediction of the future post-process measurement.