Multi-task Recurrent Neural Networks and Higher-order Markov Random Fields for Stock Price Movement Prediction: Multi-task RNN and Higer-order MRFs for Stock Price Classification.
Chang LiDongjin SongDacheng TaoPublished in: KDD (2019)
Keyphrases
- feature extraction
- multi task
- stock price
- markov random field
- recurrent neural networks
- higher order
- feature selection
- stock market
- learning tasks
- belief propagation
- non stationary
- image classification
- high order
- piecewise constant functions
- pairwise
- parameter estimation
- multi class
- graph cuts
- random fields
- conditional random fields
- feature vectors
- historical data
- neural network
- potential functions
- gaussian processes
- image restoration
- feature space
- image segmentation
- feed forward
- maximum a posteriori
- higher order cliques
- transfer learning
- energy function
- learning problems
- artificial neural networks
- discriminative random fields
- message passing
- model selection
- natural images
- support vector machine
- machine learning
- supervised learning
- markov networks
- training set
- loopy belief propagation
- support vector