No abstract

Matching a raw GPS trajectory to roads on a digital map is often referred to as the Map Matching problem. However, the occurrence of the low-sampling-rate trajectories (e.g. one point per 2 minutes) has brought lots of challenges to existing map matching algorithms. To address this problem, we propose an Interactive Voting-based Map Matching (IVMM) algorithm based on the following three insights: 1) The position context of a GPS point as well as the topological information of road networks, 2) the mutual influence between GPS points (i.e., the matching result of a point references the positions of its neighbors; in turn, when matching its neighbors, the position of this point will also be referenced), and 3) the strength of the mutual influence weighted by the distance between GPS points (i.e., the farther distance is the weaker influence exists). In this approach, we do not only consider the spatial and temporal information of a GPS trajectory but also devise a voting-based strategy to model the weighted mutual influences between GPS points. We evaluate our IVMM algorithm based on a userlabeled real trajectory dataset. As a result, the IVMM algorithm outperforms the related method (ST-Matching algorithm).

Macromolecules incorporating a highly branched polystyrene core and a poly(ethylene oxide) shell were synthesized. A comb-branched (generation G = 0) polystyrene was prepared by initiating the polymerization of styrene with sec-butyllithium, capping with 1,1-diphenylethylene, and titrating the living anions with a solution of chloromethylated linear polystyrene. A twice-grafted (G = 1) core with protected hydroxyl end groups was obtained using (6-lithiohexyl)acetaldehyde acetal to initiate the polymerization of styrene, followed by capping and grafting on the chloromethylated comb polymer. The acetal functionalities were hydrolyzed, and the core was titrated in solution with potassium naphthalide, before adding ethylene oxide. To maintain a narrow apparent molecular weight distribution, it was necessary to eliminate residual chloromethyl sites by a metal−halogen exchange reaction, prior to shell growth. Core-shell polymers based on a G = 1 core with M̄ w = 7 × 105 g·mol-1 containing 19% and 66% poly(ethylene oxide) by weight were prepared, with apparent polydispersities M̄ w/M̄ n ≈ 1.1−1.2. Another sample incorporating a G = 4 core with M̄ w of ∼108 g·mol-1 containing 36% poly(ethylene oxide) by weight was also synthesized. The hydrodynamic radii of the core and core-shell polymers were determined by dynamic light scattering. Based on the M̄ w estimated for the poly(ethylene oxide) chains, the hydrophilic chains exist in a randomly coiled conformation. The solubility behavior of the macromolecules is consistent with a core-shell morphology:  the amphiphilic copolymers are easily desolvated from tetrahydrofuran solutions, giving transparent dispersions in water or methanol.

No abstract

Combinatorial features are essential for the success of many commercial models. Manually crafting these features usually comes with high cost due to the variety, volume and velocity of raw data in web-scale systems. Factorization based models, which measure interactions in terms of vector product, can learn patterns of combinatorial features automatically and generalize to unseen features as well. With the great success of deep neural networks (DNNs) in various fields, recently researchers have proposed several DNNbased factorization model to learn both low-and high-order feature interactions. Despite the powerful ability of learning an arbitrary function from data, plain DNNs generate feature interactions implicitly and at the bit-wise level. In this paper, we propose a novel Compressed Interaction Network (CIN), which aims to generate feature interactions in an explicit fashion and at the vector-wise level. We show that the CIN share some functionalities with convolutional neural networks (CNNs) and recurrent neural networks (RNNs). We further combine a CIN and a classical DNN into one unified model, and named this new model eXtreme Deep Factorization Machine (xDeepFM). On one hand, the xDeepFM is able to learn certain bounded-degree feature interactions explicitly; on the other hand, it can learn arbitrary low-and high-order feature interactions implicitly. We conduct comprehensive experiments on three real-world datasets. Our results demonstrate that xDeepFM outperforms state-of-the-art models. We have released the source code of xDeepFM at https:// github.com/ Leavingseason/ xDeepFM.