Forecast of complex financial big data using model tree optimized by bilevel evolution strategy

If a decision tree is constructed through a series of locally optimal solutions, such as the Greedy method, overfitting to the data is likely to occur. In order to avoid overfitting, many previous research have attempted to collectively optimize the structure of a decision tree by using evolutionary computation. However, if attributes of each split and their thresholds are searched simultaneously, the evaluation function becomes intermittent; thus, optimization methods assuming continuous distribution cannot be used. In this study, in order to enable efficient search assuming continuous distribution even for complicated data that contains a lot of noise and extraordinary values, such as financial time series data, the inner level search that optimizes each threshold value collectively given a specific attribute for each split in a model tree and the outer level search that optimizes the attributes of each split were performed by separate evolutionary computing. As a result, we obtained high prediction accuracy that far exceeded the performance of the conventional method.

Hierarchical Fuzzy Systems Integrated with Particle Swarm Optimization for Daily Reference Evapotranspiration Prediction: A Novel Approach

Reference evapotranspiration (ET0) is a crucial element for deriving a meaningful scheduling of irrigation for major crops. Thus, precise projection of future ET0 is essential for better management of scarce water resources in many parts of the globe. This study evaluates the potential of a Hierarchical Fuzzy System (HFS) optimized by Particle Swarm Optimization (PSO) algorithm (PSO-HFS) to predict daily ET0. The meteorological variables and estimated ET0 were employed as inputs and outputs, respectively, for the PSO-HFS model. The FAO 56 PM method to ET0 computation was implemented to obtain ET0 values using the climatic variables obtained from two weather stations located in Gazipur Sadar and Ishurdi, Bangladesh. Prediction accuracy of PSO-HFS was compared with that of a FIS, M5 Model Tree, and a Regression Tree (RT) model. Several statistical performance evaluation indices were used to evaluate the performances of the PSO-HFS, FIS, M5 Model Tree, and RT in estimating daily ET0. Ranking of the models was performed using the concept of Shannon’s Entropy that accounts for a set of performance evaluation indices. Results revealed that the PSO-HFS model performed better than the tree-based models. Generalization capabilities of the preposed models were evaluated using the dataset from a test station (Ishurdi station). Results revealed that the models performed equally well with the unseen test dataset, and that the PSO-HFS model provided superior performance over other tree based models. The overall results imply that PSO-HFS model could effectively be utilized to model ET0 values quite efficiently and accurately.

Forecast of Complex Financial Big Data Using Model Tree Optimized By Bilevel Evolution Strategy

If a decision tree is constructed through a series of locally optimal solutions, such as the Greedy method, overfitting to the data is likely to occur. In order to avoid overfitting, many previous research have attempted to collectively optimize the structure of a decision tree by using evolutionary computation. However, if attributes of each split and their thresholds are searched simultaneously, the evaluation function becomes intermittent; thus, optimization methods assuming continuous distribution cannot be used. In this study, in order to enable efficient search assuming continuous distribution even for complicated data that contains a lot of noise and extraordinary values, such as financial time series data, the inner level search that optimizes each threshold value collectively given a specific attribute for each split in a model tree and the outer level search that optimizes the attributes of each split were performed by separate evolutionary computing. As a result, we obtained high prediction accuracy that far exceeded the performance of the conventional method.

Optimized hybrid decoupled visual servoing with supervised learning

This study proposes an optimized hybrid visual servoing approach to overcome the imperfections of classical two-dimensional, three-dimensional and hybrid visual servoing methods. These imperfections are mostly convergence issues, non-optimized trajectories, expensive calculations and singularities. The proposed method provides more efficient optimized trajectories with shorter camera path for the robot than image-based and classical hybrid visual servoing methods. Moreover, it is less likely to lose the object from the camera field of view, and it is more robust to camera calibration than the classical position-based and hybrid visual servoing methods. The drawbacks in two-dimensional visual servoing are mostly related to the camera retreat and rotational motions. To tackle these drawbacks, rotations and translations in Z-axis have been separately controlled from three-dimensional estimation of the visual features. The pseudo-inverse of the proposed interaction matrix is approximated by a neuro-fuzzy neural network called local linear model tree. Using local linear model tree, the controller avoids the singularities and ill-conditioning of the proposed interaction matrix and makes it robust to image noises and camera parameters. The proposed method has been compared with classical image-based, position-based and hybrid visual servoing methods, both in simulation and in the real world using a 7-degree-of-freedom arm robot.

Application of M5 Model Tree in Passive Remote Sensing of Thin Ice Cloud Microphysical Properties in Terahertz Region

This article presents a new method for retrieving the Ice Water Path (IWP), the median volume equivalent sphere diameter (Dme) of thin ice clouds (IWP < 100 g/m2, Dme < 80 μm) in the Terahertz band. The upwelling brightness temperature depressions caused by the ice clouds at 325.15, 448.0, 664.0 and 874.0 GHz channels are simulated by the Atmospheric Radiative Transfer Simulator (ARTS). The simulated forward radiative transfer models are taken as historical data for the M5 model tree algorithm to construct a set of piecewise functions which represent the relation of simulated brightness temperature depressions and IWP. The inversion results are optimized by an empirical relation of the IWP and the Dme for thin ice clouds which is summarized by previous studies. We inverse IWP and Dme with the simulated brightness temperature and analyze the inversion performance of selected channels. The 874.4 ± 6.0 GHz channel provides the most accurate results, because of the strong brightness temperature response to the change of IWP in the forward radiative transfer model. In order to improve the thin ice clouds IWP and Dme retrieval accuracy at the middle-high frequency channels in Terahertz band, a dual-channel inversion method was proposed that combines the 448.0 ± 3.0 GHz and 664.0 ± 4.2 GHz channel. The error analysis shows that the results of the 874.4 ± 6.0 GHz channel and the dual-channel inversion are reliable, and the IWP inversion results meet the error requirement range proposed by previous studies.