Search

Deep deterministic policy gradient (DDPG)-based car-following strategy can break through the constraints of the differential equation model due to the ability of exploration on complex environments. However, the car-following performance of DDPG is usually degraded by unreasonable reward function design, insufficient training, and low sampling efficiency. In order to solve this kind of problem, a hybrid car-following strategy based on DDPG and cooperative adaptive cruise control (CACC) is proposed. First, the car-following process is modeled as the Markov decision process to calculate CACC and DDPG simultaneously at each frame. Given a current state, two actions are obtained from CACC and DDPG, respectively. Then, an optimal action, corresponding to the one offering a larger reward, is chosen as the output of the hybrid strategy. Meanwhile, a rule is designed to ensure that the change rate of acceleration is smaller than the desired value. Therefore, the proposed strategy not only guarantees the basic performance of car-following through CACC but also makes full use of the advantages of exploration on complex environments via DDPG. Finally, simulation results show that the car-following performance of the proposed strategy is improved compared with that of DDPG and CACC., Comment: 9 pages, 11 figures

Motivation: Cryo-Electron Tomography (cryo-ET) visualizes structure and spatial organization of macromolecules and their interactions with other subcellular components inside single cells in the close-to-native state at sub-molecular resolution. Such information is critical for the accurate understanding of cellular processes. However, subtomogram classification remains one of the major challenges for the systematic recognition and recovery of the macromolecule structures in cryo-ET because of imaging limits and data quantity. Recently, deep learning has significantly improved the throughput and accuracy of large-scale subtomogram classification. However often it is difficult to get enough high-quality annotated subtomogram data for supervised training due to the enormous expense of labeling. To tackle this problem, it is beneficial to utilize another already annotated dataset to assist the training process. However, due to the discrepancy of image intensity distribution between source domain and target domain, the model trained on subtomograms in source domainmay perform poorly in predicting subtomogram classes in the target domain. Results: In this paper, we adapt a few shot domain adaptation method for deep learning based cross-domain subtomogram classification. The essential idea of our method consists of two parts: 1) take full advantage of the distribution of plentiful unlabeled target domain data, and 2) exploit the correlation between the whole source domain dataset and few labeled target domain data. Experiments conducted on simulated and real datasets show that our method achieves significant improvement on cross domain subtomogram classification compared with baseline methods., Comment: This article has been accepted for publication in Bioinformatics Published by Oxford University Press

Density estimation is a fundamental problem in both statistics and machine learning. In this study, we proposed Roundtrip as a general-purpose neural density estimator based on deep generative models. Roundtrip retains the generative power of generative adversarial networks (GANs) but also provides estimates of density values. Unlike previous neural density estimators that put stringent conditions on the transformation from the latent space to the data space, Roundtrip enables the use of much more general mappings. In a series of experiments, Roundtrip achieves state-of-the-art performance in a diverse range of density estimation tasks.

Despite the rapid development of natural language processing (NLP) implementation in electronic medical records (EMRs), Chinese EMRs processing remains challenging due to the limited corpus and specific grammatical characteristics, especially for radiology reports. In this study, we designed an NLP pipeline for the direct extraction of clinically relevant features from Chinese radiology reports, which is the first key step in computer-aided radiologic diagnosis. The pipeline was comprised of named entity recognition, synonyms normalization, and relationship extraction to finally derive the radiological features composed of one or more terms. In named entity recognition, we incorporated lexicon into deep learning model bidirectional long short-term memory-conditional random field (BiLSTM-CRF), and the model finally achieved an F1 score of 93.00%. With the extracted radiological features, least absolute shrinkage and selection operator and machine learning methods (support vector machine, random forest, decision tree, and logistic regression) were used to build the classifiers for liver cancer prediction. For liver cancer diagnosis, random forest had the highest predictive performance in liver cancer diagnosis (F1 score 86.97%, precision 87.71%, and recall 86.25%). This work was a comprehensive NLP study focusing on Chinese radiology reports and the application of NLP in cancer risk prediction. The proposed NLP pipeline for the radiological feature extraction could be easily implemented in other kinds of Chinese clinical texts and other disease predictive tasks.

Uncertainty is inevitable in transportation system due to the stochastic change of demand and supply. It is one of the most important factors affecting travelers' choice behavior. Based on the framework of Vickrey's bottleneck model, we designed and conducted laboratory experiment to investigate the effects of stochastic bottleneck capacity on commuter departure time choice behavior. Two different scenarios with different information feedback are investigated. The experimental results show that the relationship between the mean cost (E(C)) and the standard deviation of cost (\sigma) can all be fitted approximately linearly with a positive slope \sigma=E(C)/\lambda^*-m (\lambda^*>0). This suggests that under the uncertain environment, travelers are likely to minimize their travel cost budget, defined as E(C)-\lambda^* \sigma, and \lambda^*>0 indicates that the travelers behave risk preferring. The experiments also found that providing the cost information of all departure times to the commuters lowered the commuters' risk preference coefficient (i.e., \lambda^* decreases). We propose a reinforcement learning model, which is shown to reproduce the main experimental findings well.

Conformal invariance can ameliorate or eliminate the singularities residing in the black holes, and may still exist in the strong gravity regimes close to these black holes. In this paper, we try to probe this conformal invariance by looking into the wave absorption and scattering by the nonsingular static spherical black holes. The partial and total absorption cross section, as well as the differential scattering cross section, are presented for black holes with different choices of conformal parameters. Although the photon trajectories are unchanged from the Schwarzschild case since the spacetimes are conformally related, the wave optics are affected by the conformal parameters. As a result, the absorption of waves generally increases with the conformal parameters, while the shadow of the black holes remains the same as the Schwarzschild case. Moreover, the peaks in the oscillatory pattern of scattering shift towards smaller observing angles as the conformal parameters grows, while the widths of the glory peaks do not show sensitive dependence. The unique signature of the wave absorption and scattering by the nonsingular static spherical black holes in conformal gravity thus can serve to distinguish themselves from the Schwarzschild in the low frequency regime, and from other spherical black holes of alternative gravities in the high frequency limit and glory peaks., Comment: 21 pages, 12 figures, title, abstract and some contents changed, an appendix added

Side preference is a critical behavior in conflict handling, and here the behavior is investigated with pedestrian trajectories in circle antipode experiments that own both conflicting and symmetrical participant situations. In the series of experiments, more participants(around 70\%) prefer to walk on the right side, and the statistical analyses reveal that factors such as handedness, gender, and height have no significant impacts. Further investigations show that most pedestrians actually make the side choices at the very beginning, and empirical results suggest that selecting the dominate side preference (right side in our experiments) can benefit the individual movement efficiency. To reflect the realistic side preference characteristics in simulations, a Voronoi diagram based model is introduced as well as a side preference parameter with normal distribution is formulated and calibrated. Further simulations prove that the modified model is able to reproduce realistic side preference behaviors in circle antipode experiments and other common situations., Comment: 20 pages, 11 figures