A5102768296- Stock Market Forecasting Methods
- Financial Markets and Investment Strategies
- Complex Systems and Time Series Analysis
- Data Quality and Management
- Advanced Neural Network Applications
- Advanced Measurement and Detection Methods
- Electronic Health Records Systems
- Artificial Intelligence in Healthcare
- Music and Audio Processing
- Time Series Analysis and Forecasting
- Market Dynamics and Volatility
- Remote-Sensing Image Classification
- Domain Adaptation and Few-Shot Learning
- Advanced Vision and Imaging
- Simulation and Modeling Applications
Virginia Tech
2023-2024
Remote sensing image semantic segmentation is an important problem for remote interpretation. Although remarkable progress has been achieved, existing deep neural network methods suffer from the reliance on massive training data. Few-shot aims at learning to segment target objects a query using only few annotated support images of class. Most few-shot stem primarily their sole focus extracting information images, thereby failing effectively address large variance in appearance and scales...
We present DC-Gaussian, a new method for generating novel views from in-vehicle dash cam videos. While neural rendering techniques have made significant strides in driving scenarios, existing methods are primarily designed videos collected by autonomous vehicles. However, these limited both quantity and diversity compared to videos, which more widely used across various types of vehicles capture broader range scenarios. Dash often suffer severe obstructions such as reflections occlusions on...
Stock volatility prediction is an important task in the financial industry. Recent advancements multimodal methodologies, which integrate both textual and auditory data, have demonstrated significant improvements this domain, such as earnings calls (Earnings are public available often involve management team of a company interested parties to discuss company's earnings). However, these methods faced two drawbacks. First, they fail yield reliable models overfit data due their absorption...
The stock price prediction task holds a significant role in the financial domain and has been studied for long time. Recently, large language models (LLMs) have brought new ways to improve these predictions. While recent (FinLLMs) shown considerable progress NLP tasks compared smaller pre-trained (PLMs), challenges persist forecasting. Firstly, effectively integrating modalities of time series data natural fully leverage capabilities remains complex. Secondly, FinLLMs focus more on analysis...
For both investors and policymakers, forecasting the stock market is essential as it serves an indicator of economic well-being. To this end, we harness power social media data, a rich source public sentiment, to enhance accuracy predictions. Diverging from conventional methods, pioneer approach that integrates sentiment analysis, macroeconomic indicators, search engine historical prices within multi-attention deep learning model, masterfully decoding complex patterns inherent in data. We...
Predicting stock market is vital for investors and policymakers, acting as a barometer of the economic health. We leverage social media data, potent source public sentiment, in tandem with macroeconomic indicators government-compiled statistics, to refine predictions. However, prior research using tweet data prediction faces three challenges. First, quality tweets varies widely. While many are filled noise irrelevant details, only few genuinely mirror actual scenario. Second, solely focusing...
Predicting stock market is vital for investors and policymakers, acting as a barometer of the economic health. We leverage social media data, potent source public sentiment, in tandem with macroeconomic indicators government-compiled statistics, to refine predictions. However, prior research using tweet data prediction faces three challenges. First, quality tweets varies widely. While many are filled noise irrelevant details, only few genuinely mirror actual scenario. Second, solely focusing...