Traffic prediction is of great importance to traffic management and public safety, and very challenging as it is affected by many complex factors, such as spatial dependency of complicated road networks and temporal dynamics, and many more. The factors make traffic prediction a challenging task due to the uncertainty and complexity of traffic states. In the literature, many research works have applied deep learning methods on traffic prediction problems combining convolutional neural networks (CNNs) with recurrent neural networks (RNNs), which CNNs are utilized for spatial dependency and RNNs for temporal dynamics. However, such combinations cannot capture the connectivity and globality of traffic networks. In this paper, we first propose to adopt residual recurrent graph neural networks (Res-RGNN) that can capture graph-based spatial dependencies and temporal dynamics jointly. Due to gradient vanishing, RNNs are hard to capture periodic temporal correlations. Hence, we further propose a novel hop scheme into Res-RGNN to utilize the periodic temporal dependencies. Based on Res-RGNN and hop Res-RGNN, we finally propose a novel end-to-end multiple Res-RGNNs framework, referred to as “MRes-RGNN”, for traffic prediction. Experimental results on two traffic datasets have demonstrated that the proposed MRes-RGNN outperforms state-of-the-art methods significantly.

Traffic flow prediction is crucial for public safety and traffic management, and remains a big challenge because of many complicated factors, e.g., multiple spatio-temporal dependencies, holidays, and weather. Some work leveraged 2D convolutional neural networks (CNNs) and long short-term memory networks (LSTMs) to explore spatial relations and temporal relations, respectively, which outperformed the classical approaches. However, it is hard for these work to model spatio-temporal relations jointly. To tackle this, some studies utilized LSTMs to connect high-level layers of CNNs, but left the spatio-temporal correlations not fully exploited in low-level layers. In this work, we propose novel spatio-temporal CNNs to extract spatio-temporal features simultaneously from low-level to high-level layers, and propose a novel gated scheme to control the spatio-temporal features that should be propagated through the hierarchy of layers. Based on these, we propose an end-to-end framework, multiple gated spatio-temporal CNNs (MGSTC), for citywide traffic flow prediction. MGSTC can explore multiple spatio-temporal dependencies through multiple gated spatio-temporal CNN branches, and combine the spatio-temporal features with external factors dynamically. Extensive experiments on two real traffic datasets demonstrates that MGSTC outperforms other state-of-the-art baselines.

No abstract

Nonlinear regression has been extensively employed in many computer vision problems (e.g., crowd counting, age estimation, affective computing). Under the umbrella of deep learning, two common solutions exist i) transforming nonlinear regression to a robust loss function which is jointly optimizable with the deep convolutional network, and ii) utilizing ensemble of deep networks. Although some improved performance is achieved, the former may be lacking due to the intrinsic limitation of choosing a single hypothesis and the latter usually suffers from much larger computational complexity. To cope with those issues, we propose to regress via an efficient "divide and conquer" manner. The core of our approach is the generalization of negative correlation learning that has been shown, both theoretically and empirically, to work well for non-deep regression problems. Without extra parameters, the proposed method controls the bias-variance-covariance trade-off systematically and usually yields a deep regression ensemble where each base model is both "accurate" and "diversified." Moreover, we show that each sub-problem in the proposed method has less Rademacher Complexity and thus is easier to optimize. Extensive experiments on several diverse and challenging tasks including crowd counting, personality analysis, age estimation, and image super-resolution demonstrate the superiority over challenging baselines as well as the versatility of the proposed method.

Benefiting from convenient cycling and flexible parking locations, the Dockless Public Bicycle-sharing (DL-PBS) network becomes increasingly popular in many countries. However, redundant and low-utility stations waste public urban space and maintenance costs of DL-PBS vendors. In this article, we propose a Bicycle Station Dynamic Planning (BSDP) system to dynamically provide the optimal bicycle station layout for the DL-PBS network. The BSDP system contains four modules: bicycle drop-off location clustering, bicycle-station graph modeling, bicycle-station location prediction, and bicycle-station layout recommendation. In the bicycle drop-off location clustering module, candidate bicycle stations are clustered from each spatio-temporal subset of the large-scale cycling trajectory records. In the bicycle-station graph modeling module, a weighted digraph model is built based on the clustering results and inferior stations with low station revenue and utility are filtered. Then, graph models across time periods are combined to create a graph sequence model. In the bicycle-station location prediction module, the GGNN model is used to train the graph sequence data and dynamically predict bicycle stations in the next period. In the bicycle-station layout recommendation module, the predicted bicycle stations are fine-tuned according to the government urban management plan, which ensures that the recommended station layout is conducive to city management, vendor revenue, and user convenience. Experiments on actual DL-PBS networks verify the effectiveness, accuracy, and feasibility of the proposed BSDP system.

To evaluate safety and efficacy of one-vs. two-session radiofrequency ablation (RFA) of parathyroid hyperplasia for patients with secondary hyperparathyroidism (SHPT) and to compare the outcome of both methods on hypocalcemia. Patients with secondary hyperparathyroidism underwent ultrasound guided RFA of parathyroid hyperplasia. Patients were alternately assigned to either group 1 (n = 28) with RFA of all 4 glands in one session or group 2 (n = 28) with RFA of 2 glands in a first session and other 2 glands in a second session. Serum parathyroid hormone (PTH), calcium, phosphorus and alkaline phosphatase (ALP) values were measured at a series of time points after RFA. RFA parameters, including operation duration and ablation time and hospitalization length and cost, were compared between the two groups. Mean PTH decreased in group 1 from 1865.18 ± 828.93 pg/ ml to 145.72 ± 119.27 pg/ml at 1 day after RFA and in group 2 from 2256.64 ± 1021.72 pg/ml to 1388.13 ± 890.15 pg/ml at 1 day after first RFA and to 137.26 ± 107.12 pg/ml at 1 day after second RFA. Group 1's calcium level decreased to 1.79 ± 0.31 mmol/L at day 1 after RFA and group 2 decreased to 1.89 ± 0.26 mmol/L at day 1 after second session RFA (P < 0.05). Multivariate analysis showed that hypocalcemia was related to serum ALP. Patients with ALP ≥ 566 U/L had lower calcium compared to patients with ALP < 566 U/L up to a month after RFA (P < 0.05). Group 1's RFA time and hospitalization were shorter and had lower cost compared with Group 2. US-guided RFA of parathyroid hyperplasia is a safe and effective method for treating secondary hyperparathyroidism. Single-session RFA was more cost-effective and resulted in a shorter hospital stay compared to two sessions. However, patients with two-session RFA had less hypocalcemia, especially those with high ALP.Secondary hyperparathyroidism (SHPT) commonly occurs in patients with end stage renal disease (ESRD) when low calcium levels and high phosphorus levels over time stimulate increased PTH secretion 1-3 . SHPT increases the risk for osteoporosis and kidney stones, as well as for parathyroid hyperplasia, a condition that can cause mental abnormalities, renal osteodystrophy, calcific uremic arteriolopathy, vascular calcification, muscle spasms and even lead to respiratory or cardiac arrest 4 .Treatment for SHPT includes vitamin D sterols, intravenous vitamin D analogs and cinacalcet 5-10 to improve biochemical profiles and other surrogate markers 11 . Patients with severe SHPT may be candidates for parathyroidectomy (PTX), which increases long-term survival and reduces the risk of fracture in ESRD patients 12 . However, hyperparathyroidism recurs in up to 30% of patients treated with PTX due to incomplete excision of all hyperplasic parathyroid glands 13,14 . Meanwhile, PTX can potentially result in permanent hypoparathyroidism if the parathyroid glands are over-excised 15 .

No abstract

No abstract