Narrow Results

Based on the daily price data of Shanghai and London gold spot markets, we applied detrended cross-correlation analysis (DCCA) and detrended moving average cross-correlation analysis (DMCA) methods to quantify power-law cross-correlation between domestic and international gold markets. Results show that the cross-correlations between the Chinese domestic and international gold spot markets are multifractal. Furthermore, forward DMCA and backward DMCA seems to outperform DCCA and centered DMCA for short-range gold series, which confirms the comparison results of short-range artificial data in L. Y. He and S. P. Chen [Physica A 390 (2011) 3806–3814]. Finally, we analyzed the local multifractal characteristics of the cross-correlation between Chinese domestic and international gold markets. We show that multifractal characteristics of the cross-correlation between the Chinese domestic and international gold markets are time-varying and that multifractal characteristics were strengthened by the financial crisis in 2007–2008.

Under the realistic background that the capital market nowadays is a fractal market, this paper embeds the detrended cross-correlation analysis (DCCA) into the return-risk criterion to construct a Mean-DCCA portfolio model, and gives an analytical solution. Based on this, the validity of Mean-DCCA portfolio model is verified by empirical analysis. Compared to the mean-variance portfolio model, the Mean-DCCA portfolio model is more conducive for investors to build a sophisticated investment portfolio under multi-time-scale, improve the performance of portfolios, and overcome the defect that the mean-variance portfolio model has not considered the existence of fractal correlation characteristics between assets.

This study proposes a novel approach to investigating the multifractality of time series using the multifractal cross-correlation detrended moving average analysis (MF-X-DMA). The study demonstrates the behavioral differences of MF-X-DMA in coherent and non-coherent time periods. Due to the lack of a mechanism to capture the dynamical cross-correlation in time series, correlated time series with multifractal structure present a barrier for analysis. The study shows that when the wavelet coherence method is applied to time series, co-movement between time series can be easily captured in certain time intervals, providing an efficient way to find time intervals to apply MF-X-DMA. The study applies the wavelet coherence method to the daily spot prices of gold and platinum from January 1987. It shows that the wavelet coherence method is an excellent engine to extract designated time series in certain frequency and time intervals, eliminating the need for windowing or shuffling methods. Additionally, the study observes a long-term power law cross-correlation using detrended cross-correlation analysis coefficients of inversed series for both low-correlated and high-correlated series. Finally, the findings indicate that MF-X-DMA leads to superior results compared to MF-DFA when provided with highly correlated data.

Accurate identification of Golgi protein types can provide useful clues to reveal the correlation between GA dysfunction and disease pathology and improve the ability to develop more effective treatments for the diseases. This paper introduces an effective and robust method to classify Golgi protein type with traditional machine learning algorithms. In which various features such as n-GDip, DCCA, psePSSM were used as training features and SVM with linear kernel was employed as a classifier. To solve the imbalance problem of the benchmark datasets, the oversampling technique SMOTE was adopted. To deal with the huge amount of features, the PCA algorithm and Fisher feature selection method were adopted to reduce feature dimensions and remove redundant features. The experimental results show that the proposed method had a further improvement compared with other traditional machine learning methods in 10-fold cross-validation, Jackknife cross-validation and independent testing, which means a further step for the clinical application of computational methods to predict the Golgi protein types.