Joint Regression and Linear Combination of Time Series for Optimal Prediction.
Dries GeebelenKim BatselierPhilippe DreesenMarco SignorettoJohan A. K. SuykensBart De MoorJoos VandewallePublished in: ESANN (2012)
Keyphrases
- linear combination
- locally linear
- basis functions
- prediction accuracy
- small number
- regression model
- prediction model
- sparse coding
- dynamic programming
- financial time series
- multiple kernel learning
- low rank
- weighted sum
- basis vectors
- optimal solution
- orthogonal basis
- multi task
- high dimensional data
- multi variate
- set of basis functions