ISSN 0253-2778

CN 34-1054/N

2018 Vol. 48, No. 11

Display Method:
Original Paper
Outlier detection of Yangtze River basin meteorological databased on robust S-estimator
JIN Baisuo, LI Chikun
2018, 48(11): 869-876. doi: 10.3969/j.issn.0253-2778.2018.11.001
Outlier is unavoidable in high-dimensional data, such as meteorological data, and the the most widely used least-square method has no robustness and sensitivity in detecting outliers. Robust estimation can make the estimators not strongly influenced by outliers, so that the outliers can be better identified. By adding Tukey’s biweight function constraints, a principal component analysis model based on robust S-estimator was established, which converges rapidly and does not need to assume the specific form of the distribution function. Then the observations were smoothed by B-spline basis, the mean residuals squared norm was used as the test statistic, and the adjusted box-plot which also has robustness was trained to detect the outliers. In the example, more than 58 thousand measurements of meteorological data over 60 years of 5 cities in Yangtze River basin were adopted. A comparative analysis of the data set with outlier detecting procedure based on principal component analysis and robust S-estimator has been conducted. It can be seen clearly that compared with the classical approach, the outlier detecting procedure based on robust S-estimator gives more information on the abnormal data, and thus can identify outliers better.
The threshold dividends in the Cramr-Lundberg risk model with loss-carry forward tax payments
LIU Zhang, MA Yue, HU Yijun, XIAO Liqun
2018, 48(11): 877-884. doi: 10.3969/j.issn.0253-2778.2018.11.002
A classical Cramr-Lundberg risk model with a threshold dividend strategy and loss-carry forward tax payments was studied. For this model, the closed-form expressions for the expected accumulated discounted dividends until ultimate ruin was derived and the explicit solution was presented when the individual claim amount follows an exponential distribution. Finally, numerical
On a class of locally dually flat (α,β)-metrics
HUA Yiping, SONG Weidong
2018, 48(11): 885-889. doi: 10.3969/j.issn.0253-2778.2018.11.003
Locally dually flat weak Landsberg(α,β)-metrics in the form of F=α(βα) were studied,
On self-dual and LCD double circulant codes over Fq+uFq+vFq+uvFq
LU Yaqi, SHI Minjia, WU Wenting, XIAO Aqing
2018, 48(11): 890-897. doi: 10.3969/j.issn.0253-2778.2018.11.004
Double circulant codes of length 2n over a non-chain ring Fq+uFq+vFq+uvFq, u2=v2=0, uv=vu, were studied when q was a prime power. Exact enumerations of self-dual and LCD double circulant codes for a positive integer n were given. Using a distance-preserving Gray map, self-dual and LCD codes of length 8n over Fq were constructed when q was even. Using random coding and the Artin conjecture, the modified Varshamov-Gilbert bounds were derived on the relative distance of the codes considered, building on exact enumeration results for given n and q.
Complemented Sylow subgroups of finite groups
BAO Hongwei, ZHANG Jia, MIAO Long
2018, 48(11): 898-901. doi: 10.3969/j.issn.0253-2778.2018.11.005
The relationship between the complementability of Sylow subgroups and the structure of a group was considered, and some results about the construction of composition factors were obtained. Further, one of Heliel’s results is the corollary of our results.
Planar order on vertex poset
LU Xuexing
2018, 48(11): 902-905. doi: 10.3969/j.issn.0253-2778.2018.11.006
A planar order is a special linear extension of the edge poset (partially ordered set) of a processive plane graph. The definition of a planar order makes sense for any finite poset and is equivalent to the one of a conjugate order. Here it was proved that there is a planar order on the vertex poset of a processive planar graph naturally induced from the planar order of its edge poset.
Pricing and hedging barrier options based on Merton model and Monte Carlo simulation
ZHENG Xiang, WEI Yongfeng
2018, 48(11): 906-922. doi: 10.3969/j.issn.0253-2778.2018.11.007
Barrier options as typical exotic options are trading frequently at the domestic OTC (over-the counter) market, whose jumping structure and path dependence make the hedge method a constant problem for the industry. Here a barrier options hedge strategy applicable to current domestic financial market was designed by a comparative analysis for pricing the up-and-out barrier options of the CSI 300 index. The barrier options price and the Greeks change were analyzed through the analytical solution of the Black-Scholes-Merton model and numerical solution of the Monte Carlo simulation method. According to the simulated 10 000 index path of hedge and the variation of delta, the static replicate maximal cost was enumerated and options dynamic barrier out-shift boundary was deduced for analyzing the average hedge cost and extreme effect. 2011~2016 actual CSI 300 was selected to back-test and verify the effectiveness of the hedge strategy. The results show that the hedge average cost is significantly reduced under the hedge idea of the traversal trigger outward moving barrier boundary, and that the distribution of the hedge extremes and quantiles is relatively smooth, which reflects the good performance of the hedge strategy and the effective hedge of the barrier option.
High-frequency trading strategies based on deep learning algorithms and their profitability
SUN Dachang, BI Xiuchung
2018, 48(11): 923-932. doi: 10.3969/j.issn.0253-2778.2018.11.008
As an important algorithm, deep learning has been applied successfully to image processing, speech recognition, machine translation and other fields. Here, deep learning algorithms were applied to high-frequency trading. Convolutional neural network(CNN) and long short-term memory(LSTM) neural network were selected to build up and down classification models, respectively. Based on the models, high-frequency trading strategies were proposed. Then the data of bitumen futures contract was used for back-testing and empirically analyzing the superiority of the strategies. In back-testing, deep learning algorithms were compared with artificial neural network(ANN). The results show that both strategies based on CNN and LSTM neural network exhibit better profitability and generalization ability. In addition, the winning rates and expected returns of the two strategies are also better.
On the improvement of closed-loop supply chain systems based on brittle theory
SHAO Ruirui, FANG Zhigeng, XU Xinwei, LIU Sifeng, TAO Liangyan, NIE Yuanyuan
2018, 48(11): 933-942. doi: 10.3969/j.issn.0253-2778.2018.11.009
For the closed-loop supply chain system consisting of manufacturers, retailers and demand markets, entropy theory was used to quantify the brittleness of the closed-loop supply chain system as a whole, and it was studied that how, with the overall degree of brittleness of the closed-loop supply chain system exceeding the threshold, resulting in the collapse of the system, to weaken the brittle relationship between systems while reducing the overall brittleness of the system, so as to maintain the stability of the system. By analyzing the brittle structure and characteristics of the closed-loop supply chain system, the brittle propagation model was established, the hierarchical brittleness diagram of the system was obtained, and the brittle source of the closed-loop supply chain system was identified. The brittleness of the system was reduced from three aspects: the brittle source, the source of the brittle source, and the subsystem with the largest correlation with the brittle source. Through examples, it is shown that when the overall brittleness of the closed-loop supply chain system exceeds the system threshold, in order to avoid system crashes, the brittleness of the system is reduced through the treatment of brittle sources, thus improving the system.
Design and experiment of twin plate device based on electrowetting-on-dielectric actuation
XU Xiaowei, ZHANG Yuliang
2018, 48(11): 943-948. doi: 10.3969/j.issn.0253-2778.2018.11.010
Digital microfluidics based on electrowetting on dielectric is an emerging popular technology that manipulates single droplets. It has shown enormous advantages in biology, medicine and chemistry and so on, where it has been used extensively. However, the higher driving voltage of digital microfluidic devices not only causes the dielectric layer of the chip to be broken down, but the strong electric field can cause irreversible damage to the active material in the droplet. Therefore, it is necessary to reduce the driving voltage of the digital microfluidic device. Compared with the two plate structure, the twin plate device composed of two identical coplanar electrodes was obtained by theoretical analysis, which can not only achieve greater driving force, but also lower the threshold driving voltage. The experimental results demonstrate that the twin plate structure can improve the average velocity of droplets and reduce the driving voltage effectively. Especially, in the lower driving voltage a better droplet driving effect can be obtained with the twin plate structure.
Transdermal delivery peptide promoted the regulation of mice hair growth by epidermal growth factor
JIN Peipei, WANG Yanshi, ZHOU Xin, YE Shulai, WANG Ying, WAN Xin
2018, 48(11): 949-954. doi: 10.3969/j.issn.0253-2778.2018.11.011
TD1 was known as a peptide chaperon with 11 amino acids (ACSSSPSKHCG) that can facilitate transdermal delivery of proteins. Previous studies showed the maximum transdermal efficiency of fusion protein TD1-hEGF(T-E) was 16 folds higher than that of native hEGF. The second-generation of TD1-hEGF fused protein: TD1-hEGF-TD1(T-E-T) and TD1-TD1-hEGF (2T-E), designed based on the fact that transdermal efficiency was concentration-dependent of TD1, possessed considerable higher transdermal abilities than TD1-hEGF. To further detect the bioactivity of transdermal hEGF mediated by TD1, a hair growth experiments in mice was conducted. It was found that all the three fused proteins above can induce the transition of hair follicles from the telogen phase to the anagen phase of the hair cycle. These findings provide convincing evidences for improved practicability of the transdermal delivery peptide TD1 in cosmetic.