Search

Representation-based methods have found widespread applications in various classification tasks. However, these methods cannot deal effectively with imbalanced data scenarios. They tend to neglect the importance of minority samples, resulting in bias toward the majority class. To address this limitation, we propose a density-based discriminative nonnegative representation approach for imbalanced classification tasks. First, a new class-specific regularization term is incorporated into the framework of a nonnegative representation based classifier (NRC) to reduce the correlation between classes and improve the discrimination ability of the NRC. Second, a weight matrix is generated based on the hybrid density information of each sample’s neighbors and the decision boundary, which can assign larger weights to minority samples and thus reduce the preference for the majority class. Furthermore, the resulting model can be efficiently optimized through the alternating direction method of multipliers. Extensive experimental results demonstrate that our proposed method is superior to numerous state-of-the-art imbalanced learning methods. [ABSTRACT FROM AUTHOR]

Collaborative representation-based classification (CRC) has been extensively applied to various recognition fields due to its effectiveness and efficiency. Nevertheless, it is generally suboptimal for imbalanced learning. Previous studies have revealed that a class-imbalance distribution can lead CRC, and even most conventional classification methods, to ignore the minority class and prioritize the majority class. To address this deficiency, this paper proposes a hybrid density-based adaptive weighted collaborative representation model that incorporates a regularization technique and an adaptive weight generation mechanism into the CRC framework. A new regularization term, based on class-specific representation, is introduced to decrease the correlation between classes and enhance CRC's discriminative ability. The sample distribution and density information within and between classes are employed to assign greater weights to minority samples, thereby strengthening the representation capabilities of minority samples and reducing the bias towards the majority class. Furthermore, it is theoretically demonstrated that this model has a closed-form solution. Its complexity is comparable to that of CRC, ensuring its efficiency. Extensive experiments on diverse data sets from the KEEL repository show the superiority of the proposed method compared to other state-of-the-art imbalanced classification methods. [ABSTRACT FROM AUTHOR]

Railway track settlements are extremely important for the safe operation of trains. This study utilizes an in situ test and a three-dimensional finite element method (3D FEM) to investigate the settlement of underpasses and railway tracks caused by shield tunneling. First, the 3D FEM results are compared with field monitoring data. Second, the settlement variation of tracks during shield construction is analyzed. Finally, the effect of the tunnel–underpass distance is observed. A comparative study establishes the reliability of the achieved 3D FEM results, which showed the consistency of railway track settlement with that of the underpass top plate. Based on the cutter head location, the underpass and railway track settlements were divided into three stages. With an increase in the tunnel–underpass distance, the underpass and railway track settlements increased, and the speed of increase was significantly higher when the tunnel was beyond the scope of the underpass. When the tunnel axis was within the scope of the underpass, the settlement of the bottom plate > top plate > railway track, and the differences between them were relatively small. Contrarily, the top plate settlement was larger than the bottom plate, while the railway track settlement increased. In both cases, the difference between the top and bottom plate settlements gradually decreased as the tunnel–underpass distance increased. This study shows that the underpass may have a settlement control effect on the deformation of railway tracks of high stiffness. [ABSTRACT FROM AUTHOR]

The damage variation of a masonry structure during shield tunneling has been investigated. Furthermore, the damage degree of the masonry building has been investigated by combining the field measurement, finite element method results, and theoretical method. The results show that compared with the theoretical calculation method LSTM, the method provide by this paper can give the detail damage location of the masonry structure caused by shield construction. Due to the existence of door and window openings, the result of JMM is larger than LSTM result, the differences can be modified by concept of "characteristic tensile strain". The wall of the masonry building encounter the shield face first suffers more damage than the later ones. At the beginning of tunnlling, the damage were generated from a small area at the bottom for wall 1, but a larger area at the top of the building for wall 2. The damage area increase more at the top of the building as tunneling advanced, but the maximum damage occurred at the bottom of the building. [ABSTRACT FROM AUTHOR]

A soil pressure balance rectangular pipe jacking project is investigated to find out the interaction of jacking control and ground settlement in Zhengzhou, China. The dimension of the rectangular pipe jacking excavation face is 6.94 m (width) × 4.24 m (height). The tunnel is 35.6 m long and 10.6 m beneath the ground in silty clay strata. The trajectory of the pipe jacking tunneling machine was measured during the whole construction. Meanwhile the construction sets like cutter pressure, jacking force, and grouting pressure were measured with sensors on the pipe jacking machine. The ground settlement was obtained throughout the rectangular pipe jacking process. The results showed that ground settlement variation started from 7.46 m in front of the cutter and ended with 9.79 m away from cutter. The ground settlement can be divided into four phases according to the settlement increasing pattern which is consistent with the distance from the cutter. The ratio of these four phases settlement is approximately 3:3:2:2. The trajectory of the jacking pipe tail has a nonnegligible effect on the settlement distribution characteristics of the ground. The vertical trajectory variation induces lager settlement increasing than horizontal trajectory. Ground settlement increases during the whole progress, but the increasing amplitude decreases with the increasing of cutter pressure, jacking force and grouting. [ABSTRACT FROM AUTHOR]

Load-bearing pile foundations induce soil stress changes and ground deformation, which affect the safety of nearby tunnels. Centrifuge model tests and numerical simulations were conducted to investigate the effects of bored piles on existing tunnels in clay. The effects of pile length, pile–tunnel distance, and pile–tunnel depth on the tunnel moment and settlement were analyzed by comparing the results of both centrifuge model tests and numerical simulations. The results revealed that the tunnel longitudinal moment increased with the pile load but decreased with the net distance between the pile and tunnel. The circumferential moments near the pile bottom were larger than those near the other parts. The maximum tunnel moments were observed for the case with the pile and tunnel at the same depth. The tunnel moments for the case with the pile tip above the tunnel were greater than those in the case with the pile tip below the tunnel. The results obtained in this study provide valuable insight into the effect of loaded piles on nearby existing tunnels and the design of buildings near subways. [ABSTRACT FROM AUTHOR]

Due to the low computational complexity, the two-phase test sample representation method shows outstanding advantages in the field of real-time face recognition. However, its first phase does not fully consider the imbalance of the determined K-nearest training samples. This defect is not well reflected and will seriously affect the recognition accuracy of the second phase. In this paper, we explore the above issues and propose to incorporate the unselected training samples into the modeling process. The proposed method not only allows the unselected samples to perform the final representation but also makes the selected nearest training samples play a more significant role than the unselected ones in classification. It is efficiently solved with an analytical solution. The rationales and probability interpretation of the proposed method are presented to further guarantee its rationality and effectiveness. Extensive experiments on diverse face databases are conducted to demonstrate the superior recognition performance in comparison with other competing classifiers. [ABSTRACT FROM AUTHOR]

Background: Colloid-facilitated phosphorus (P) transport is recognized as an important pathway for the loss of soil P in agricultural systems; however, information regarding soil aggregate-associated colloidal P (Pcoll) is lacking. To elucidate the effects of aggregate size on the potential loss of Pcoll in agricultural systems, soils (0–20 cm depth) from six land-use types were sampled in the Zhejiang Province in the Yangtze River Delta region, China. The aggregate size fractions (2–8 mm, 0.26–2 mm, 0.053–0.26 mm and < 0.053 mm) were separated using the wet sieving method. Colloidal P and other soil parameters in aggregates were analyzed. Results: Our study demonstrated that 0.26–2 mm small macroaggregates had the highest total P (TP) content. In acidic soils, the highest Pcoll content was observed in the 0.26- to 2-mm-sized aggregates, while the lowest was reported in the < 0.053 mm (silt + clay)-sized particles, the opposite of that revealed in alkaline and neutral soils. Paddy soils contained less Pcoll than other land-use types. The proportion of Pcoll in total dissolved P (TDP) was dominated by < 0.053 mm (silt + clay)-sized particles. Aggregate size strongly influenced the loss potential of Pcoll in paddy soils, where Pcoll contributed up to 83% TDP in the silt + clay-sized particles. The Pcoll content was positively correlated with TP, Al, Fe, and the mean weight diameter. Aggregate-associated total carbon (TC), total nitrogen (TN), C/P, and C/N had significant negative effects on the contribution of Pcoll to potential soil P loss. The Pcoll content of the aggregates was controlled by the aggregate-associated TP and Al content, as well as the soil pH value. The potential loss of Pcoll from aggregates was controlled by its organic matter content. Conclusion: We concluded that management practices that increase soil aggregate stability or its organic carbon content will limit Pcoll loss in agricultural systems. [ABSTRACT FROM AUTHOR]

In this paper, the main factors impacting the plug flow pattern of a clearwell were investigated by integrating pilot-scale, full-scale clearwell tracer testing and computational fluid dynamics (CFD) simulation. It was found that pilot tracer testing, full-scale tracer testing and CFD simulation all demonstrated that the correlation between the ratio of t 10/ T and L/W can be approximately expressed by: t 10/ T = 0.189 4ln( L/W)-0.049 4. This study confirmed that the installation of baffles within clearwells is an efficient way to optimize their configuration. In addition, the inlet velocity has a minimal contribution to the ratio of t 10/ T. However, the ratio of turning channel width to channel width ( d/W) significantly contributes to the ratio of t 10/ T. The optimal ratio of d/W is 0.8–1.2 for maintaining better plug flow pattern. The number of turning channels is one of the main factors that impact the ratio of t 10/ T. When increasing the number of turning channels, a lower ratio of t 10/ T is obtained. [ABSTRACT FROM AUTHOR]