Seeking an appropriate alternative least squares algorithm for nonnegative tensor factorizations - A novel recursive solution for nonnegative quadratic programming and NTF.
Anh Huy PhanAndrzej CichockiPublished in: Neural Comput. Appl. (2012)
Keyphrases
- quadratic programming
- least squares
- linear programming
- tensor factorization
- objective function
- optimal solution
- computational complexity
- nonnegative matrix factorization
- parameter estimation
- k means
- dynamic programming
- learning algorithm
- closed form
- expectation maximization
- np hard
- support vector machine svm
- clustering algorithm
- probabilistic model
- prior information
- ls svm
- image processing
- training set