GCN2CDD A Commercial District Discovery Framework via Embedding Space Clustering on Graph Convolution Networks
ABSTARCT :
Modern enterprises attach much attention to the selection of commercial locations. With the rapid development of urban data and machine learning, we can discover the patterns of human mobility with these data and technology to guide commercial district discovery. In this paper, we propose an unsupervised commercial district discovery framework via embedding space clustering on graph convolution networks (GCN2CDD) to solve the problem of commercial district discovery. Specifically, the proposed framework aggregates human mobility features according to geographic similarity by graph convolution networks. Based on the graph convolution networks embedding space, we apply hierarchical clustering to mine the latent functional regions hidden in different human patterns. Then with the kernel density estimation, we can obtain the semantic labels for the clustering results to discover commercial districts. Finally, we analyze the multi-sources data of the Xiaoshan District and Chengdu City and experiments verify the effectiveness of our framework.
EXISTING SYSTEM :
? We provide a detailed review over existing graph neural network models. We present a general design pipeline and discuss the variants of each module. We also introduce researches on theoretical and empirical analyses of GNN models.
? Edge-level tasks are edge classification and link prediction, which require the model to classify edge types or predict whether there is an edge existing between two given nodes.
? It uses the confidence-driven scheme to adaptively select the starting node and determine the node updating sequence.
? It follows the same idea of generalizing the existing LSTMs into the graph-structured data but has a specific updating sequence while methods mentioned above are agnostic to the order of nodes.
DISADVANTAGE :
? Many deep learning models have been proposed to solve the traffic flow prediction problem with the more expansive urban road sensing device arrangement and improved recognition accuracy.
? Therefore, the Cross-correlation method overcomes the phase sliding problem and compares the shape similarity of two-time series.
? In practice, the normalized value of Cross-correlation is usually used to limit the range to be within [-1, 1], where 1 means strong correlation and -1 means that they are completely opposite.
? Traffic congestion on urban roads is an important issue that needs to be addressed in smart cities’ development.
PROPOSED SYSTEM :
• Based on CNNs and graph embedding, variants of graph neural networks (GNNs) are proposed to collectively aggregate information from graph structure.
• Apart from different variants of spatial approaches, several general frameworks are proposed aiming to integrate different models into one single framework.
• Neural network-based models require abundant labeled data and it is costly to obtain enormous human-labeled data.
• Selfsupervised methods are proposed to guide models to learn from unlabeled data which is easy to obtain from websites or knowledge bases.
ADVANTAGE :
? We used the Simulation of Urban Mobility simulation platform to evaluate the performance of the proposed method under different traffic demand conditions, and the experimental results show that the proposed method can reduce the density of mainline bottlenecks and improve the efficiency of mainline traffic.
? We compared the performance of the proposed method and the traditional method under different traffic demands.
? The processed data is smoother and more continuous than the raw data, which is consistent with the real situation and helps the neural network’s training, and improves the neural network’s prediction performance.
? The distance-flow-based method and traffic-state-based method have a better performance on the congestion reduction.
|