Forecasting Using First-Order Difference of Time Series and Bagging of Competitive Associative Nets.
Shuichi KurogiRyohei KoyamaShinya TanakaToshihisa SanukiPublished in: IJCNN (2007)
Keyphrases
- weather forecasting
- forecasting accuracy
- financial time series
- arma model
- exponential smoothing
- chaotic time series
- box jenkins
- higher order
- ensemble methods
- short term
- first order logic
- random forest
- support vector regression
- arima model
- random forests
- neural network
- ensemble learning
- neural network model
- moving average
- stock market
- hybrid model
- non stationary
- stock price
- decision trees
- garch model
- mackey glass
- demand forecasting
- imbalanced data
- bp neural network
- dynamic time warping
- turning points
- machine learning
- generalization error
- associative memory
- long term
- ensemble selection
- forecasting model
- multivariate time series
- classifier ensemble
- learning machines
- autoregressive
- randomized trees
- voting methods
- prediction model
- prediction accuracy
- autoregressive integrated moving average