Credit Risk Prediction Model for Listed Companies Based on CNN-LSTM and Attention Mechanism

Electronics, Journal Year: 2023, Volume and Issue: 12(7), P. 1643 - 1643

Published: March 30, 2023

The financial market has been developing rapidly in recent years, and the issue of credit risk concerning listed companies become increasingly prominent. Therefore, predicting is an urgent concern for banks, regulators investors. commonly used models are Z-score, Logit (logistic regression model), kernel-based virtual machine (KVM) neural network models. However, results achieved could be more satisfactory. This paper proposes a credit-risk-prediction model based on CNN-LSTM attention mechanism, Our approach benefits long short-term memory (LSTM) long-term time-series prediction combined with convolutional (CNN) model. Furthermore, advantages being integrated into include reducing complexity data, improving calculation speed training solving possible lack historical data sequence LSTM model, resulting accuracy. To reduce problems, we introduced mechanism to assign weights independently optimize show that our distinct compared other CNNs, LSTMs, CNN-LSTMs research credit-risk listing formula significant meaning.

Language: Английский

---

Multi-Object Multi-Camera Tracking Based on Deep Learning for Intelligent Transportation: A Review

Sensors, Journal Year: 2023, Volume and Issue: 23(8), P. 3852 - 3852

Published: April 10, 2023

Multi-Objective Multi-Camera Tracking (MOMCT) is aimed at locating and identifying multiple objects from video captured by cameras. With the advancement of technology in recent years, it has received a lot attention researchers applications such as intelligent transportation, public safety self-driving driving technology. As result, large number excellent research results have emerged field MOMCT. To facilitate rapid development need to keep abreast latest current challenges related field. Therefore, this paper provide comprehensive review multi-object multi-camera tracking based on deep learning for transportation. Specifically, we first introduce main object detectors MOMCT detail. Secondly, give an in-depth analysis evaluate advanced methods through visualisation. Thirdly, summarize popular benchmark data sets metrics quantitative comparisons. Finally, point out faced transportation present practical suggestions future direction.

Language: Английский

---

Real-time load forecasting model for the smart grid using bayesian optimized CNN-BiLSTM

Frontiers in Energy Research, Journal Year: 2023, Volume and Issue: 11

Published: May 5, 2023

A smart grid is a new type of power system based on modern information technology, which utilises advanced communication, computing and control technologies employs sensors, measurement, communication devices that can monitor the status operation various in real-time optimise dispatch through intelligent algorithms to achieve efficient system. However, due its complexity uncertainty, how effectively perform prediction an important challenge. This paper proposes model attention mechanism convolutional neural network (CNN) combined with bi-directional long short-term memory BiLSTM.The has stronger spatiotemporal feature extraction capability, more accurate capability better adaptability than ARMA decision trees. The traditional models tree often only use simple statistical methods for prediction, cannot meet requirements high accuracy efficiency load so CNN-BiLSTM Bayesian optimisation following advantages suitable compared tree. CNN hierarchical structure containing several layers such as layer, pooling layer fully connected layer. mainly used extracting features from data images, dimensionality reduction features, classification recognition. core operation, locally weighted summation input extract data. In convolution different be extracted by setting kernels BiLSTM capture semantic dependencies both directions. consists two LSTM process sequence forward backward directions combine obtain comprehensive contextual information. access front back inputs at each time step results. It prevents gradient explosion disappearance while capturing longer-distance dependencies. extracts then optimises them Bayes. By collecting system, including power, load, weather other factors, our uses deeply learn grids key future prediction. Meanwhile, algorithm model’s hyperparameters, thus improving performance. provide reference help energy utilisation

Language: Английский

---

Automatic Lip-reading Classification Using Deep Learning Approaches and Optimized Quaternion Meixner Moments by GWO algorithm

Knowledge-Based Systems, Journal Year: 2024, Volume and Issue: 304, P. 112430 - 112430

Published: Sept. 5, 2024

Language: Английский

---

RL-GCN: Traffic flow prediction based on graph convolution and reinforcement learning for smart cities

Displays, Journal Year: 2023, Volume and Issue: 80, P. 102513 - 102513

Published: Sept. 4, 2023

Language: Английский

---

An incremental learning approach to dynamic parallel machine scheduling with sequence-dependent setups and machine eligibility restrictions

Applied Soft Computing, Journal Year: 2024, Volume and Issue: 164, P. 112002 - 112002

Published: July 15, 2024

Language: Английский

---

Causal embedding of user interest and conformity for long-tail session-based recommendations

Information Sciences, Journal Year: 2023, Volume and Issue: 644, P. 119167 - 119167

Published: May 18, 2023

Language: Английский

---

A WOA-CNN-BiLSTM-based multi-feature classification prediction model for smart grid financial markets

Frontiers in Energy Research, Journal Year: 2023, Volume and Issue: 11

Published: May 16, 2023

Introduction: Smart grid financial market forecasting is an important topic in deep learning. The traditional LSTM network widely used time series because of its ability to model and forecast data. However, long-term forecasting, the lack historical data may lead a decline performance. This difficult problem for networks overcome. Methods: In this paper, we propose new deep-learning address problem. WOA-CNN-BiLSTM combines bidirectional long short-term memory BiLSTM convolution Advantages Neural Network CNN. We replace with network, BiLSTM, exploit capturing dependencies. It can capture dependencies modelling. At same time, use convolutional neural (CNN) extract features better represent patterns regularity method combining CNN learn characteristics more comprehensively, thus improving accuracy prediction. Then,to further improve performance CNN-BiLSTM model, optimize using whale algorithm WOA. optimization algorithm, which has good global search convergence speed, complete short time. Results: Optimizing through WOA reduce calculation training prediction smart market, market. Experimental results show that our proposed than other models effectively deal missing sequence forecasting. Discussion: provides necessary help development markets risk management services, promote growth industry. Our research are great significance learning, provide effective idea solving grid.

Language: Английский

---

MS3Net: Multiscale stratified-split symmetric network with quadra-view attention for hyperspectral image classification

Signal Processing, Journal Year: 2023, Volume and Issue: 212, P. 109153 - 109153

Published: June 15, 2023

Language: Английский

---

Intelligent large-scale flue-cured tobacco grading based on deep densely convolutional network

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: July 10, 2023

Abstract Flue-cured tobacco grading plays a crucial role in leaf purchase and the formulation of groups. However, traditional flue-cured mode is usually manual, which time-consuming, laborious, subjective. Hence, it essential to research more efficient intelligent methods. Most existing methods suffer from classes less accuracy problem. Meanwhile, limited by different industry applications, datasets are hard be obtained publicly. The employ relatively small lower resolution data that apply practice. Therefore, aiming at insufficiency feature extraction ability inadaptability multiple grades, we collected largest highest dataset proposed an method based on deep densely convolutional network (DenseNet). Diverging other approaches, our has unique connectivity pattern neural concatenates preceding data. This connects all previous layers subsequent layer directly for transmission. idea can better extract depth image information features transmit each layer’s data, thereby reducing loss encouraging reuse. Then, designed whole pre-processing process experimented with learning algorithms verify usability. experimental results showed DenseNet could easily adapted changing output fully connected layers. With 0.997, significantly higher than methods, came best model solving

Language: Английский

---