A5100362104- Advanced Neural Network Applications
- Financial Markets and Investment Strategies
- Economic theories and models
- Stochastic processes and financial applications
- Stock Market Forecasting Methods
- Complex Systems and Time Series Analysis
- Credit Risk and Financial Regulations
- Banking stability, regulation, efficiency
- Multimodal Machine Learning Applications
- Video Surveillance and Tracking Methods
- Brain Tumor Detection and Classification
- Industrial Vision Systems and Defect Detection
- Domain Adaptation and Few-Shot Learning
- Advanced Image and Video Retrieval Techniques
- Rough Sets and Fuzzy Logic
- Semiconductor Quantum Structures and Devices
- Adversarial Robustness in Machine Learning
- Monetary Policy and Economic Impact
- Radiomics and Machine Learning in Medical Imaging
- Service-Oriented Architecture and Web Services
- Explainable Artificial Intelligence (XAI)
- Advanced Data and IoT Technologies
- Fuzzy Systems and Optimization
- Human Pose and Action Recognition
- Housing Market and Economics
Nanchang Hangkong University
2023-2025
Northeastern University
2021-2025
Boston University
2025
Shenyang Ligong University
2024
Xiangtan University
2023-2024
Beijing Normal University
2024
China Electronics Technology Group Corporation
2023
National University of Defense Technology
2023
Luoyang Institute of Science and Technology
2023
Guangdong Polytechnic Normal University
2023
Major events often trigger abrupt changes in stock prices and volatility. We study the implications of jumps volatility on investment strategies. Using event‐risk framework Duffie, Pan, Singleton (2000) , we provide analytical solutions to optimal portfolio problem. Event risk dramatically affects strategy. An investor facing event is less willing take leveraged or short positions. The acts as if some portion his wealth may become illiquid strategy blends both dynamic buy‐and‐hold Jumps have...
We derive the optimal investment policy of a risk-averse investor in market where there is textbook arbitrage opportunity, but liabilities must be secured by collateral. find that it often to underinvest taking smaller position than collateral constraints allow. Even when followed, portfolio typically experiences losses before final convergence date. In fact, its initial performance may indistinguishable from conventional with poor track record. These results have important implications for...
We study how the market prices default and liquidity risks incorporated into one of most important credit spreads in financial markets-interest rate swap spreads.Our approach consists jointly modeling Treasury, repo, term structures using a general five-factor affine framework estimating parameters by maximum likelihood.We find that spread is driven changes persistent process rapidly mean-reverting intensity process.Although both processes have similar volatilities, we premium priced rates...
We examine the efficiency of using individual stocks or portfolios as base assets to test asset pricing models cross-sectional data. The literature has argued that creating reduces idiosyncratic volatility and allows more precise estimates factor loadings, consequently risk premia. show analytically empirically smaller standard errors portfolio beta do not lead coefficient estimates. Factor premia are determined by distributions loadings residual risk. Portfolios destroy information...
Osteosarcoma is a malignant tumor derived from primitive osteogenic mesenchymal cells, which extremely harmful to the human body and has high mortality rate. Early diagnosis treatment of this disease necessary improve survival rate patients, MRI an effective tool for detecting osteosarcoma. However, due complex structure variable location osteosarcoma, cancer cells are highly heterogeneous prone aggregation overlap, making it easy doctors inaccurately predict area lesion. In addition, in...
Fine-tuning helps large language models (LLM) recover degraded information and enhance task performance.Although Low-Rank Adaptation (LoRA) is widely used effective for fine-tuning, we have observed that its scaling factor can limit or even reduce performance as the rank size increases. To address this issue, propose RoRA (Rank-adaptive Reliability Optimization), a simple yet method optimizing LoRA's factor. By replacing $\alpha/r$ with $\alpha/\sqrt{r}$, ensures improved Moreover, enhances...
Recently Xia and Song [Phys. Lett. A 364 (2007) 117] have proposed a controlled quantum secure direct communication (CQSDC) protocol. They claimed that in their protocol only with the help of controller Charlie, receiver Alice can successfully extract secret message from sender Bob. In this letter, first we will show within Charlie's role could been excluded if it were not for unreasonable design. We then revise Xia–Song CQSDC such its original advantages are retained be really realized.
The ReLU/Leaky-ReLU activation functions are widely used in convolutional neural network (CNN) architectures due to their simple implementation and outstanding performance. However, the current lack spatial priors among pixels during feature extraction. As a result, they may struggle capture critical piecewise semantic features that essential for image segmentation. Drawing inspiration from regional characteristics of human brain neurons, this paper presents novel spatially regularized...
Lotus pods in unstructured environments often present multi-scale characteristics the captured images. As a result, it makes their automatic identification difficult and prone to missed false detections. This study proposed lightweight lotus pod model, MLP-YOLOv5, deal with this difficulty. The model adjusted detection layer optimized anchor box parameters enhance small object accuracy. C3 module transformer encoder (C3-TR) shuffle attention (SA) mechanism were introduced improve feature...
The main focus of this paper is to investigate the multiple attribute decision making (MADM) method under intuitionistic linguistic (IL) environment, based on induced aggregation operators and analyze possibilities for its application in low carbon supplier selection. More specifically, a new operator, called weighted ordered averaging (ILWIOWA), introduced facilitate IL information. Some desired properties are explored. A further generalization ILWIOWA, generalized (ILGWIOWA), operator...
While many studies document that the market risk premium is predictable and betas are not constant, dividend discount model ignores time-varying premiums betas.We develop a to consistently value cashflows with changing risk-free rates, conditional in context of CAPM.Practical valuation accomplished an analytic term structure different rates applied expected at horizons.Using constant can produce large mis-valuations, which, portfolio data, mostly driven short horizons by long time-variation...
An artificial stock market is established with the modeling method and ideas of cellular automata. Cells are used to represent stockholders, who have capability self-teaching affected by investing history neighboring ones. The neighborhood relationship among stockholders expanded Von Neumann relationship, interaction them realized through selection operator crossover operator. Experiment shows that large events frequent in fluctuations price generated when compared a normal process returns...
Gliomas, often known as low-grade gliomas, are malignant brain tumors. Codeletion of chromosomal arms 1p/19q has been connected with a good response to treatment in gliomas (LGG) several studies. For planning, the ability anticipate 1p19q status is crucial. This research’s purpose develop noninvasive approach based on MR images using our efficient CNNs. While public networks like VGGNet, GoogleNet, and other well-known can use transfer learning identify cancer MRI, model contains large...
The interpretability of deep learning models has emerged as a compelling area in artificial intelligence research. safety criteria for medical imaging are highly stringent, and required an explanation. However, existing convolutional neural network solutions left ventricular segmentation viewed terms inputs outputs. Thus, the CNNs come into spotlight. Since data limited, many methods to fine-tune that popular transfer have been built using massive public ImageNet datasets by method....
The recently proposed data augmentation TransMix employs attention labels to help visual transformers (ViT) achieve better robustness and performance. However, is deficient in two aspects: 1) image cropping method of may not be suitable for ViTs. 2) At the early stage training, model produces unreliable maps. uses maps compute mixed that can affect model. To address aforementioned issues, we propose MaskMix Progressive Attention Labeling (PAL) label space, respectively. In detail, from...