Optimizing AI/ML Hyperparameters with SimWrapper and OptQuest
Download our report to see how SimWrapper and OptQuest can be used to optimally tune hyperparameters for your AI/ML models
Optimizing AI/ML Hyperparameters with SimWrapper and OptQuest
An Example of How to Optimally Tune Hyperparameters in AI/ML Models
Machine learning (ML) is an application of artificial intelligence (AI) that enables models to “learn” to perform tasks using data. Some applications of ML include regression, classification, image/speech recognition, forecasting, and decision making. Using ML techniques, models are developed, or “trained”, algorithmically through exposure to data. The automated training process means that ML models do not need to be explicitly programmed; however, the parameters that define the model structure itself and control the algorithm used to train the model, known as hyperparameters, cannot be learned from the training data and need to be specified by the model developer (e.g., the number of nodes in a hidden layer of a neural network, or the learning rate of a training algorithm). These hyperparameters can fundamentally alter ML model performance, and often need to be tuned to specific values for different applications to achieve optimal performance.
Simply complete the form at the top to download our 11-page white paper.
More Optimization White Papers
Pandemic: Decision Support for a Novel Problem and Its Challenges
Closing down our economy too much to slow the spread of the virus comes with severe economic repercussions, while opening up the economy too much in an attempt to save jobs allows the virus to get a foothold. There are both political and game-theoretic pressures on how and when to open up (e.g, a region that opens early will pull economic activity from its neighbors). When policy makers face such high-stakes tradeoffs, emotion and bias inevitably creep in. We provide a mathematical and scientific approach that minimizes human bias and helps us to decide the optimal course of action can provide critical inputs to the decision process.
Tabu Search
Tabu Search, also called Adaptive Memory Programming, is a method for solving challenging optimization problems in the fields of business, engineering, economics and science.
Everyday examples include practical applications in resource management, financial and investment planning, healthcare systems, energy and environmental policy, pattern classification, biotechnology and a host of other areas.
Tabu search has emerged as one of the leading technologies for handling real-world problems that have proved difficult or impossible to solve with classical procedures.