Optimal Base Station Network Based on Topological Data Analysis

The decision of which base stations need to be removed due to the cost is always a difficult problem, because the influence on the cover rate of the network caused by the removal should be kept to a minimum. However, the common methods to solve this problem such as K-means Clustering show a low accuracy. Barcode, which belongs to TDA, has the possibility to show the result by identifying the Persistent Homology of base station network. This essay mainly illustrates the specific problem of optimal base station network, which applies the TDA(Topological Data Analysis) methods to find which base stations need removing due to the cost K-means Clustering and Topological Data Analysis methods were mainly used. With the simulated distribution of telecommunication users, K-means Clustering algorithm was used to locate 30 best base stations. By comparing the minimum distance between the results (K=25 and K=30), K-means Clustering was used again to decide base station points to be removed. Then TDA was used to select which 5 base stations should be removed through observing barcode. By repeating above steps five times, Finally the average and variance of cover area in original network, K-means Clustering and TDA were compared. The experiment showed that the average cover rate of original network was 81.20% while the result of TDA and K-means Clustering were 92.13% and 89.87%. It was proved by simulation that it is more efficient to use TDA methods to construct the optimal base station network.

Merging Arcs to Produce Acyclic Phylogenetic Networks and Normal Networks

As phylogenetic networks grow increasingly complicated, systematic methods for simplifying them to reveal properties will become more useful. This paper considers how to modify acyclic phylogenetic networks into other acyclic networks by contracting specific arcs that include a set D. The networks need not be binary, so vertices in the networks may have more than two parents and/or more than two children. In general, in order to make the resulting network acyclic, additional arcs not in D must also be contracted. This paper shows how to choose D so that the resulting acyclic network is “pre-normal”. As a result, removal of all redundant arcs yields a normal network. The set D can be selected based only on the geometry of the network, giving a well-defined normal phylogenetic network depending only on the given network. There are CSD maps relating most of the networks. The resulting network can be visualized as a “wired lift” in the original network, which appears as the original network with each arc drawn in one of three ways.

SMYOLO: Lightweight Pedestrian Target Detection Algorithm in Low-Altitude Scenarios

Due to the arbitrariness of the drone’s shooting angle of view and camera movement and the limited computing power of the drone platform, pedestrian detection in the drone scene poses a greater challenge. This paper proposes a new convolutional neural network structure, SMYOLO, which achieves the balance of accuracy and speed from three aspects: (1) By combining deep separable convolution and point convolution and replacing the activation function, the calculation amount and parameters of the original network are reduced; (2) by adding a batch normalization (BN) layer, SMYOLO accelerates the convergence and improves the generalization ability; and (3) through scale matching, reduces the feature loss of the original network. Compared with the original network model, SMYOLO reduces the accuracy of the model by only 4.36%, the model size is reduced by 76.90%, the inference speed is increased by 43.29%, and the detection target is accelerated by 33.33%, achieving minimization of the network model volume while ensuring the detection accuracy of the model.

Analysis of semi-open queueing networks using lost customers approximation with an application to robotic mobile fulfilment systems

We consider a semi-open queueing network (SOQN), where one resource from a resource pool is needed to serve a customer. If on arrival of a customer some resource is available, the resource is forwarded to an inner network to complete the customer’s order. If no resource is available, the new customer waits in an external queue until one becomes available (“backordering”). When a resource exits the inner network, it is returned to the resource pool. We develop a new solution approach. In a first step we modify the system such that new arrivals are lost if the resource pool is empty (“lost customers”). We adjust the arrival rate of the modified system such that the throughputs in all nodes of the inner network are pairwise identical to those in the original network. Using queueing theoretical methods, in a second step we reduce this inner network to a two-station system including the resource pool. For this two-station systems, we invert the first step and obtain a standard SOQN which can be solved analytically. We apply our results to storage and delivering systems with robotic mobile fulfilment systems (RMFSs). Instead of sending pickers to the storage area to search for the ordered items and pick them, robots carry shelves with ordered items from the storage area to picking stations. We model the RMFS as an SOQN to determine the minimal number of robots.

The Impacts of Surface Microchannels on the Transport Properties of Porous Fibrous Media Using Stochastic Pore Network Modeling

A stochastic pore network modeling method with tailored structures is proposed to investigate the impacts of surface microchannels on the transport properties of porous fibrous media. Firstly, we simplify the original pore network extracted from the 3D images. Secondly, a repeat sampling strategy is applied during the stochastic modeling of the porous structure at the macroscale while honoring the structural property of the original network. Thirdly, the microchannel is added as a spherical chain and replaces the overlapped elements of the original network. Finally, we verify our model via a comparison of the structure and flow properties. The results show that the microchannel increases the permeability of flow both in the directions parallel and vertical to the microchannel direction. The microchannel plays as the highway for the pass of reactants while the rest of the smaller pore size provides higher resistance for better catalyst support, and the propagation path in the network with microchannels is more even and predictable. This work indicates that our modeling framework is a promising methodology for the design optimization of cross-scale porous structures.

Facial expression recognition based on improved VGG convolutional neural network

With the development of artificial intelligence, facial expression recognition based on deep learning has become a current research hotspot. The article analyzes and improves the VGG16 network. First, the three fully connected layers of the original network are changed to two convolutional layers and one fully connected layer, which reduces the complexity of the network; Then change the maximum pooling in the network to local-based adaptive pooling to help the network select feature information that is more conducive to facial expression recognition, so that the network can be used on the facial expression datasets RAF-DB and SFEW. The recognition rate increased by 4.7% and 7% respectively.

Deep Convolutional Neural Network Optimization for Defect Detection in Fabric Inspection

This research is aimed to detect defects on the surface of the fabric and deep learning model optimization. Since defect detection cannot effectively solve the fabric with complex background by image processing, this research uses deep learning to identify defects. However, the current network architecture mainly focuses on natural images rather than the defect detection. As a result, the network architecture used for defect detection has more redundant neurons, which reduces the inference speed. In order to solve the above problems, we propose network pruning with the Bayesian optimization algorithm to automatically tune the network pruning parameters, and then retrain the network after pruning. The training and detection process uses the above-mentioned pruning network to predict the defect feature map, and then uses the image processing flow proposed in this research for the final judgment during fabric defect detection. The proposed method is verified in the two self-made datasets and the two public datasets. In the part of the proposed network optimization results, the Intersection over Union (IoU) of four datasets are dropped by 1.26%, 1.13%, 1.21%, and 2.15% compared to the original network model, but the inference time is reduced to 20.84%, 40.52%, 23.02%, and 23.33% of the original network model using Geforce 2080 Ti. Furthermore, the inference time is also reduced to 17.56%, 37.03%, 19.67%, and 22.26% using the embedded system AGX Xavier. After the image processing part, the accuracy of the four datasets can reach 92.75%, 94.87%, 95.6%, and 81.82%, respectively. In this research, Yolov4 is also trained with fabric defects, and the results showed this model are not conducive to detecting long and narrow fabric defects.

Predictive Sampling Method for Spread Models in Networks

This paper proposes a novel, exploration-based network sampling algorithm called caterpillar quota walk sampling (CQWS) inspired by the caterpillar tree. Network sampling identifies a subset of nodes and edges from a network, creating an induced graph. Beginning from an initial node, exploration-based sampling algorithms grow the induced set by traversing and tracking unvisited neighboring nodes from the original network. Tunable and trainable parameters allow CQWS to maximize the sum of the degrees of the induced graph from multiple trials when sampling dense networks. A network spread model renders effective use in various applications, including tracking the spread of epidemics, visualizing information transmissions through social media, and cell-to-cell spread of neurodegenerative diseases. CQWS generates a spread model as its sample by visiting the highest-degree neighbors of previously visited nodes. For each previously visited node, a top proportion of the highest-degree neighbors fulfills a quota and branches into a new caterpillar tree. Sampling more high-degree nodes constitutes an objective among various applications. Many exploration-based sampling algorithms suffer drawbacks that limit the sum of degrees of visited nodes and thus the number of high-degree nodes visited. Furthermore, a strategy may not be adaptable to volatile degree frequencies throughout the original network architecture, which influences how deep into the original network an algorithm could sample. This paper analyzes CQWS in comparison to four other exploration-based network in tackling these two problems by sampling sparse and dense randomly generated networks.

Optimizing the Simplicial-Map Neural Network Architecture

Simplicial-map neural networks are a recent neural network architecture induced by simplicial maps defined between simplicial complexes. It has been proved that simplicial-map neural networks are universal approximators and that they can be refined to be robust to adversarial attacks. In this paper, the refinement toward robustness is optimized by reducing the number of simplices (i.e., nodes) needed. We have shown experimentally that such a refined neural network is equivalent to the original network as a classification tool but requires much less storage.

Probabilistic network sparsification with ego betweenness

Sparsification is the process of decreasing the number of edges in a network while one or more topological properties are preserved. For probabilistic networks, sparsification has only been studied to preserve the expected degree of the nodes. In this work we introduce a sparsification method to preserve ego betweenness. Moreover, we study the effect of backboning and density on the resulting sparsified networks. Our experimental results show that the sparsification of high density networks can be used to efficiently and accurately estimate measures from the original network, with the choice of backboning algorithm only partially affecting the result.