APPLYING MONTE CARLO AND MACHINE LEARNING METHODS FOR ROBUST CONVEX OPTIMIZATION BASED PREDICTION ALGORITHMS

摘要

Disclosed herein are methods and systems for improving accuracy of convex optimization-based prediction models while reducing their computation load by clustering a plurality of received random variables extracted from a plurality of samples to a plurality of clusters based on their covariance matrix, applying a convex optimization based prediction model to compute predicted optimal intra-cluster solutions for each cluster and an optimal inter-cluster solution over the plurality of clusters and predicting optimal solutions for the plurality of samples by based on an aggregation between the optimal intra-cluster solutions and the optimal inter-cluster solution. Separately computing the intra-cluster and the inter-cluster solution reduces the prediction model's computation load. Further disclosed are methods and systems for selecting a best performing prediction model for a certain dataset using Monte Carlo algorithms to generate simulated samples based on the received samples and computing estimation errors for the prediction models applied to the simulated samples.