scholarly journals Two-Stage Label Embedding via Neural Factorization Machine for Multi-Label Classification

2019 ◽  
Vol 33 ◽  
pp. 3304-3311
Author(s):  
Chen Chen ◽  
Haobo Wang ◽  
Weiwei Liu ◽  
Xingyuan Zhao ◽  
Tianlei Hu ◽  
...  
Keyword(s):  
Learning Model ◽  
Two Stage ◽  
Second Stage ◽  
Proposed Model ◽  
Latent Space ◽  
Label Correlations ◽  
Factorization Machine ◽  
Classification Tasks ◽  
Real World Datasets ◽  
Embedding Methods

Label embedding has been widely used as a method to exploit label dependency with dimension reduction in multilabel classification tasks. However, existing embedding methods intend to extract label correlations directly, and thus they might be easily trapped by complex label hierarchies. To tackle this issue, we propose a novel Two-Stage Label Embedding (TSLE) paradigm that involves Neural Factorization Machine (NFM) to jointly project features and labels into a latent space. In encoding phase, we introduce a Twin Encoding Network (TEN) that digs out pairwise feature and label interactions in the first stage and then efficiently learn higherorder correlations with deep neural networks (DNNs) in the second stage. After the codewords are obtained, a set of hidden layers is applied to recover the output labels in decoding phase. Moreover, we develop a novel learning model by leveraging a max margin encoding loss and a label-correlation aware decoding loss, and we adopt the mini-batch Adam to optimize our learning model. Lastly, we also provide a kernel insight to better understand our proposed TSLE. Extensive experiments on various real-world datasets demonstrate that our proposed model significantly outperforms other state-ofthe-art approaches.

scholarly journals A Bicycle Origin–Destination Matrix Estimation Based on a Two-Stage Procedure

2020 ◽  
Vol 12 (7) ◽  
pp. 2951
Author(s):  
Seungkyu Ryu
Keyword(s):  
Gravity Model ◽  
Transportation Network ◽  
Modal Shift ◽  
Two Stage ◽  
Matrix Estimation ◽  
Second Stage ◽  
Local Mobility ◽  
Proposed Model ◽  
Path Flow Estimator

As more people choose to travel by bicycle, transportation planners are beginning to recognize the need to rethink the way they evaluate and plan transportation facilities to meet local mobility needs. A modal shift towards bicycles motivates a change in transportation planning to accommodate more bicycles. However, the current methods to estimate bicycle volumes on a transportation network are limited. The purpose of this research is to address those limitations through the development of a two-stage bicycle origin–destination (O–D) matrix estimation process that would provide a different perspective on bicycle modeling. From the first stage, a primary O–D matrix is produced by a gravity model, and the second stage refines that primary matrix generated in the first stage using a Path Flow Estimator (PFE) to build the finalized O–D demand. After a detailed description of the methodology, the paper demonstrates the capability of the proposed model for a bicycle demand matrix estimation tool with a real network case study.

scholarly journals A Two-Stage Industrial Defect Detection Framework Based on Improved-YOLOv5 and Optimized-Inception-ResnetV2 Models

2022 ◽  
Vol 12 (2) ◽  
pp. 834
Author(s):  
Zhuang Li ◽  
Xincheng Tian ◽  
Xin Liu ◽  
Yan Liu ◽  
Xiaorui Shi
Keyword(s):  
Defect Detection ◽  
Network Architecture ◽  
Feature Fusion ◽  
Visual Object ◽  
Two Stage ◽  
The Public ◽  
Second Stage ◽  
Steel Equipment ◽  
Public Dataset ◽  
Classification Tasks

Aiming to address the currently low accuracy of domestic industrial defect detection, this paper proposes a Two-Stage Industrial Defect Detection Framework based on Improved-YOLOv5 and Optimized-Inception-ResnetV2, which completes positioning and classification tasks through two specific models. In order to make the first-stage recognition more effective at locating insignificant small defects with high similarity on the steel surface, we improve YOLOv5 from the backbone network, the feature scales of the feature fusion layer, and the multiscale detection layer. In order to enable second-stage recognition to better extract defect features and achieve accurate classification, we embed the convolutional block attention module (CBAM) attention mechanism module into the Inception-ResnetV2 model, then optimize the network architecture and loss function of the accurate model. Based on the Pascal Visual Object Classes 2007 (VOC2007) dataset, the public dataset NEU-DET, and the optimized dataset Enriched-NEU-DET, we conducted multiple sets of comparative experiments on the Improved-YOLOv5 and Inception-ResnetV2. The testing results show that the improvement is obvious. In order to verify the superiority and adaptability of the two-stage framework, we first test based on the Enriched-NEU-DET dataset, and further use AUBO-i5 robot, Intel RealSense D435 camera, and other industrial steel equipment to build actual industrial scenes. In experiments, a two-stage framework achieves the best performance of 83.3% mean average precision (mAP), evaluated on the Enriched-NEU-DET dataset, and 91.0% on our built industrial defect environment.

scholarly journals Accounting for Fairness in a Two-Stage Stochastic Programming Model for Kidney Exchange Programs

2018 ◽  
Vol 15 (7) ◽  
pp. 1491 ◽  
Author(s):  
Hyunwoo Lee ◽  
Seokhyun Chung ◽  
Taesu Cheong ◽  
Sang Song
Keyword(s):  
Stochastic Programming ◽  
Programming Model ◽  
Deterministic Models ◽  
Two Stage ◽  
Total Utility ◽  
Exchange Programs ◽  
Kidney Exchange ◽  
Second Stage ◽  
Proposed Model ◽  
Stochastic Programming Model

Kidney exchange programs, which allow a potential living donor whose kidney is incompatible with his or her intended recipient to donate a kidney to another patient in return for a kidney that is compatible for their intended recipient, usually aims to maximize the number of possible kidney exchanges or the total utility of the program. However, the fairness of these exchanges is an issue that has often been ignored. In this paper, as a way to overcome the problems arising in previous studies, we take fairness to be the degree to which individual patient-donor pairs feel satisfied, rather than the extent to which the exchange increases social benefits. A kidney exchange has to occur on the basis of the value of the kidneys themselves because the process is similar to bartering. If the matched kidneys are not of the level expected by the patient-donor pairs involved, the match may break and the kidney exchange transplantation may fail. This study attempts to classify possible scenarios for such failures and incorporate these into a stochastic programming framework. We apply a two-stage stochastic programming method using total utility in the first stage and the sum of the penalties for failure in the second stage when an exceptional event occurs. Computational results are provided to demonstrate the improvement of the proposed model compared to that of previous deterministic models.

scholarly journals Incorporating Label Embedding and Feature Augmentation for Multi-Dimensional Classification

2020 ◽  
Vol 34 (04) ◽  
pp. 6178-6185 ◽  
Author(s):  
Haobo Wang ◽  
Chen Chen ◽  
Weiwei Liu ◽  
Ke Chen ◽  
Tianlei Hu ◽  
...  
Keyword(s):  
Cross Correlation ◽  
Feature Space ◽  
Dimensional Classification ◽  
Label Information ◽  
Label Correlations ◽  
Factorization Machine ◽  
Feature Augmentation ◽  
Real World Datasets ◽  
Low Dimensional ◽  
Original Feature

Feature augmentation, which manipulates the feature space by integrating the label information, is one of the most popular strategies for solving Multi-Dimensional Classification (MDC) problems. However, the vanilla feature augmentation approaches fail to consider the intra-class exclusiveness, and may achieve degenerated performance. To fill this gap, a novel neural network based model is proposed which seamlessly integrates the Label Embedding and Feature Augmentation (LEFA) techniques to learn label correlations. Specifically, based on attentional factorization machine, a cross correlation aware network is introduced to learn a low-dimensional label representation that simultaneously depicts the inter-class correlations and the intra-class exclusiveness. Then the learned latent label vector can be used to augment the original feature space. Extensive experiments on seven real-world datasets demonstrate the superiority of LEFA over state-of-the-art MDC approaches.

A new two-stage genetic programming classification algorithm and its applications

2017 ◽  
Vol 40 (8) ◽  
pp. 2560-2578
Author(s):  
Ruochen Liu ◽  
Lijia An ◽  
Xin Yu
Keyword(s):  
Genetic Programming ◽  
Selection Method ◽  
Local Optimum ◽  
Computation Complexity ◽  
Classification Problems ◽  
Two Stage ◽  
Second Stage ◽  
Image Edge Detection ◽  
Image Edge ◽  
Real World Datasets

This work aims to find a better solution from an improved genetic programming (GP) algorithm for classification problems without increasing computation costs as far as possible. Firstly, the standard GP algorithm is difficult to succeed in finding better individuals since too large search landscape and it is easy to run into local optimum when evolution reaches a certain stage, while, by enlarging the evolutionary generation or the size of population might increase computation complexity. So, a two-stage GP may be a good solution for these. In the first stage, GP is used to induce a relatively simple classifier and construct features; in the second stage, GP is executed on these features to evolve a better classifier. Secondly, in order to improve the convergence, a new initialization (NI) strategy and a new function operator selection method are designed. In this paper, a NI strategy based two-stage GP algorithm (NITGP) is proposed, and compares with the standard GP on a set of artificial, real-world datasets and image edge detection tasks. The experimental results show that our approach can evolve classifiers with better performance.

scholarly journals Urban Railway Transit Timetable Optimisation Based on Passenger-and-Trains Matching – A Case Study of Beijing Metro Line

2021 ◽  
Vol 33 (5) ◽  
pp. 671-687
Author(s):  
Junsheng Huang ◽  
Tong Zhang ◽  
Runbin Wei
Keyword(s):  
Genetic Algorithm ◽  
Waiting Time ◽  
Stage Model ◽  
Two Stage ◽  
Left Behind ◽  
Second Stage ◽  
Decision Variables ◽  
Proposed Model ◽  
Railway System

Due to the congested scenarios of the urban railway system during peak hours, passengers are often left behind on the platform. This paper firstly brings a proposal to capture passengers matching different trains. Secondly, to reduce passengers’ total waiting time, timetable optimisation is put forward based on passengers matching different trains. This is a two-stage model. In the first stage, the aim is to obtain a match between passengers and different trains from the Automatic Fare Collection (AFC) data as well as timetable parameters. In the second stage, the objective is to reduce passengers’ total waiting time, whereby the decision variables are headway and dwelling time. Due to the complexity of our proposed model, an MCMC-GASA (Markov Chain Monte Carlo-Genetic Algorithm Simulated Annealing) hybrid method is designed to solve it. A real-world case of Line 1 in Beijing metro is employed to verify the proposed two-stage model and algorithms. The results show that several improvements have been brought by the newly designed timetable. The number of unique matching passengers increased by 37.7%, and passengers’ total waiting time decreased by 15.5%.

Huber collaborative representation for robust multiclass classification

2019 ◽  
Vol 17 (04) ◽  
pp. 1950020 ◽  
Author(s):  
Cuiming Zou ◽  
Yuan Yan Tang ◽  
Yulong Wang ◽  
Zhenghua Luo
Keyword(s):  
Robust Statistics ◽  
Classification Problem ◽  
Multiclass Classification ◽  
Classification Performance ◽  
Collaborative Representation ◽  
Proposed Model ◽  
The Mean ◽  
Classification Tasks ◽  
Real World Datasets ◽  
The Cost

Recent advances have shown a great potential of collaborative representation (CR) for multiclass classification. However, conventional CR-based classification methods adopt the mean square error (MSE) criterion as the cost function, which is sensitive to gross corruption and outliers. To address this limitation, inspired by the success of robust statistics, we develop a Huber collaborative representation-based classification (HCRC) method for robust multiclass classification. Concretely, we cast the classification problem as a Huber collaborative representation problem with the Huber estimator. Our another contribution is to design an efficient half-quadratic (HQ) algorithm with guaranteed convergence to solve the proposed model efficiently. Furthermore, we also give a theoretical analysis of the classification performance of HCRC. Experiments on real-world datasets corroborate that HCRC is an effective and robust algorithm for multiclass classification tasks.

Two-Stage Distributionally Robust Optimization for a Two-Allocation p-Hub Median Problem

2021 ◽  
pp. 2150004
Author(s):  
Congke Wang ◽  
Yankui Liu ◽  
Peiyu Zhang ◽  
Guoqing Yang
Keyword(s):  
Robust Optimization ◽  
Median Problem ◽  
Two Stage ◽  
Distributionally Robust Optimization ◽  
Second Stage ◽  
Proposed Model ◽  
Hub Network ◽  
Distributionally Robust ◽  
Partial Distribution

This paper presents a novel two-stage distributionally robust optimization model of the two-allocation p-hub median problem with different hub link scales. With the objective of minimizing overall costs of building and operating the hub network, the choices of hub locations and hub link scales are decided in the first stage, while the optimal flows are determined in the second stage once the uncertain demands have been realized. Before establishing the hub network, we just have partial distribution information about the uncertain flow demands, which can be described by a given perturbation set based on the historical information. Due to the ambiguous distributions leading to a computationally intractable model, we reformulate the proposed model into the tractable robust counterpart forms under two types of uncertainty sets (Box[Formula: see text]ellipsoidal perturbation set and Generalized ellipsoidal perturbation set). Finally, to demonstrate the effectiveness and applicability for our model, we conduct a case study for the express network system in the Beijing–Tianjin–Hebei region.

scholarly journals Two-Stage Spatiotemporal Context Refinement Network for Precipitation Nowcasting

2021 ◽  
Vol 13 (21) ◽  
pp. 4285
Author(s):  
Dan Niu ◽  
Junhao Huang ◽  
Zengliang Zang ◽  
Liujia Xu ◽  
Hongshu Che ◽  
...  
Keyword(s):  
Heavy Precipitation ◽  
Machine Learning Algorithms ◽  
Two Stage ◽  
Rainfall Prediction ◽  
Multi Scale ◽  
Second Stage ◽  
Proposed Model ◽  
Radar Echo ◽  
Echo Threshold ◽  
Residual Block

Precipitation nowcasting by radar echo extrapolation using machine learning algorithms is a field worthy of further study, since rainfall prediction is essential in work and life. Current methods of predicting the radar echo images need further improvement in prediction accuracy as well as in presenting the predicted details of the radar echo images. In this paper, we propose a two-stage spatiotemporal context refinement network (2S-STRef) to predict future pixel-level radar echo maps (deterministic output) more accurately and with more distinct details. The first stage is an efficient and concise spatiotemporal prediction network, which uses the spatiotemporal RNN module embedded in an encoder and decoder structure to give a first-stage prediction. The second stage is a proposed detail refinement net, which can preserve the high-frequency detailed feature of the radar echo images by using the multi-scale feature extraction and fusion residual block. We used a real-world radar echo map dataset of South China to evaluate the proposed 2S-STRef model. The experiments showed that compared with the PredRNN++ and ConvLSTM method, our 2S-STRef model performs better on the precipitation nowcasting, as well as at the image quality evaluating index and the forecasting indices. At a given 45dBZ echo threshold (heavy precipitation) and with a 2 h lead time, the widely used CSI, HSS, and SSIM indices of the proposed 2S-STRef model are found equal to 0.195, 0.312, and 0.665, respectively. In this case, the proposed model outperforms the OpticalFlow method and PredRNN++ model.

scholarly journals Discriminative and Correlative Partial Multi-Label Learning

2019 ◽  
Author(s):  
Haobo Wang ◽  
Weiwei Liu ◽  
Yang Zhao ◽  
Chen Zhang ◽  
Tianlei Hu ◽  
...  
Keyword(s):  
Real World ◽  
State Of The Art ◽  
Feature Space ◽  
Gradient Boosting ◽  
Training Procedure ◽  
Second Stage ◽  
Label Correlations ◽  
Real World Datasets ◽  
Partial Label Learning ◽  
Confidence Value

In partial label learning (PML), each instance is associated with a candidate label set that contains multiple relevant labels and other false positive labels. The most challenging issue for the PML is that the training procedure is prone to be affected by the labeling noise. We observe that state-of-the-art PML methods are either powerless to disambiguate the correct labels from the candidate labels or incapable of extracting the label correlations sufficiently. To fill this gap, a two-stage DiscRiminative and correlAtive partial Multi-label leArning (DRAMA) algorithm is presented in this work. In the first stage, a confidence value is learned for each label by utilizing the feature manifold, which indicates how likely a label is correct. In the second stage, a gradient boosting model is induced to fit the label confidences. Specifically, to explore the label correlations, we augment the feature space by the previously elicited labels on each boosting round. Extensive experiments on various real-world datasets clearly validate the superiority of our proposed method.

Sign in / Sign up

Export Citation Format

Share Document