Moscow, Idaho, United States
Computational Mathematics and Automated Intelligence
Time-Series and Event Prediction through Neural Networks, Machine Learning and other algorithms. Unique robust predictions are created for each problem with a proprietary framework that uses large quantities of public and private information. Data can be noisy, regular or irregular, temporal or spacial. Models are rebuilt and validated for each new prediction with up to the minute information.
Developer of Neural Network Toolbox for MATLAB and consulting on machine learning problems.
Created and extended for 25 years The MathWork's core deep learning product used to train neural network models to perform diverse tasks, including function fitting, classification, clustering, time series prediction, dynamical modeling and control, and other data driven machine learning. The most significant commercial deep learning product on the planet at its release in 1992, without any significant competition until 2014. (Wikipedia, "Comparison of deep learning software")
Prediction models for customer/product demand.