Deep learning for financial applications : A survey

Deep learning for enhanced risk management: a novel approach to analyzing financial reports

Risk management is a critical component of today's financial environment because of the enormity and complexity of data contained in financial statements. Business situations, plans, and schedule risk assessment with the help of conventional ways which involve analytical, technical, and heuristic models are inadequate to address the complex structures of the latest data. This research brings out the Hybrid Financial Risk Predictor (HFRP) model, using the convolutional neural networks (CNN) and long-short term memory (LSTM) networks to improve financial risk prediction. A combination of quantitative and qualitative ratings derived from the analysis of financial texts results in high accuracy and stability compared with the HFRP model. Evaluating key findings, the quantity of training & testing loss decreased considerably and they have their final value as 0.0013 and 0.003, respectively. According to the hypothesis, the selected HFRP model demonstrates the values of the revenue, net income, and earnings per share (EPS), and are closely similar to the actual values. The model achieves substantial risk mitigation: credit risk lowered from 0.75 to 0.20, liquidity risk from 0.70 to 0.25, market risk from 0.65 to 0.30, while operational risk is at 0.80 to 0.35. By analyzing the results of the HFRP model, it can be stated that the proposal promotes improved financial stability and presents a reliable model for the contemporary financial markets, which in turn helps in making sound decisions and improve the assessment of risks.

Integrating machine learning for sustaining cybersecurity in digital banks

Cybersecurity continues to be an important concern for financial institutions given the technology's rapid development and increasing adoption of digital services. Effective safety measures must be adopted to safeguard sensitive financial data and protect clients from potential harm due to the rise in cyber threats that target digital organizations. The aim of this study is to investigates how machine learning algorithms are integrated into cyber security measures in the context of digital banking and its benefits and drawbacks. We initially provide a general overview of digital banks and the particular security concerns that differentiate them from conventional banks. Then, we explore the value of machine learning in strengthening cybersecurity defenses. We revealed that insider threats, distributed denial of service (DDoS) assaults, ransomware, phishing attacks, and social engineering are main cyberthreats that are digital banks exposed. We identify the appropriate machine learning algorithms such as support vector machines (SVM), recurrent neural networks (RNN), hidden markov models (HMM), and local outlier factor (LOF) that are used for detection and prevention cyberthreats. In addition, we provide a model that considers ethical concerns while constructing a cybersecurity framework to address potential vulnerabilities in digital banking systems. The advantages and disadvantages of incorporating machine learning into the cybersecurity strategy of digital banks are outlined using strengths, weaknesses, opportunities, threats (SWOT) analysis. This study seeks to provide a thorough knowledge of how machine learning may strengthen cybersecurity procedures, protect digital banks, and maintain customer trust in the ecosystem of digital banking.

Human-Unrecognizable Differential Private Noised Image Generation Method

Differential privacy has emerged as a practical technique for privacy-preserving deep learning. However, recent studies on privacy attacks have demonstrated vulnerabilities in the existing differential privacy implementations for deep models. While encryption-based methods offer robust security, their computational overheads are often prohibitive. To address these challenges, we propose a novel differential privacy-based image generation method. Our approach employs two distinct noise types: one makes the image unrecognizable to humans, preserving privacy during transmission, while the other maintains features essential for machine learning analysis. This allows the deep learning service to provide accurate results, without compromising data privacy. We demonstrate the feasibility of our method on the CIFAR100 dataset, which offers a realistic complexity for evaluation.

Enhancing portfolio management using artificial intelligence: literature review

Building an investment portfolio is a problem that numerous researchers have addressed for many years. The key goal has always been to balance risk and reward by optimally allocating assets such as stocks, bonds, and cash. In general, the portfolio management process is based on three steps: planning, execution, and feedback, each of which has its objectives and methods to be employed. Starting from Markowitz's mean-variance portfolio theory, different frameworks have been widely accepted, which considerably renewed how asset allocation is being solved. Recent advances in artificial intelligence provide methodological and technological capabilities to solve highly complex problems, and investment portfolio is no exception. For this reason, the paper reviews the current state-of-the-art approaches by answering the core question of how artificial intelligence is transforming portfolio management steps. Moreover, as the use of artificial intelligence in finance is challenged by transparency, fairness and explainability requirements, the case study of post-hoc explanations for asset allocation is demonstrated. Finally, we discuss recent regulatory developments in the European investment business and highlight specific aspects of this business where explainable artificial intelligence could advance transparency of the investment process.

Graphical Deep Learning Prediction Model for Stock Risk Management

Forecasting the future movements of stock market indexes by utilizing historical transaction data is a prominent concern within the realm of finance. The application of graph convolutional networks to incorporate the interrelationships among various indices’ patterns is a highly advanced subject within this field. Addressing the inconsistency between historical and future dynamic graphs in current graph convolution-based index prediction, we propose a method called G-Conv that constructs a graph structure based on constituent stocks of the indices for index trend prediction. This approach extracts traditional quantitative features along with deep features from one-dimensional convolutional networks as characteristics of prediction samples. The method produces index trend predictions by constructing a graph structure using constituent stock data of indices and applying graph convolution to different index sample features. The proposed methodology’s efficacy is verified by utilizing 42 widely employed indicators in the A-share market. The experimental findings demonstrate that when utilizing mean absolute error (MAE) and mean squared error (MSE) as the loss functions for model training, G-Conv outperforms classic methods such as GC-CNN and ADGAT. Specifically, G-Conv reduces the average prediction errors by 5.10% and 4.20% respectively, as evaluated by the two error criteria. Additionally, G-Conv exhibits favorable generalization performance.

C, Badarudeen S, Sriraman H. A systematic literature survey on recent trends in stock market prediction

Prediction of the stock market is a challenging and time-consuming process. In recent times, various research analysts and organizations have used different tools and techniques to analyze and predict stock price movements. During the early days, investors mainly depend on technical indicators and fundamental parameters for short-term and long-term predictions, whereas nowadays many researchers started adopting artificial intelligence-based methodologies to predict stock price movements. In this article, an exhaustive literature study has been carried out to understand multiple techniques employed for prediction in the field of the financial market. As part of this study, more than hundreds of research articles focused on global indices and stock prices were collected and analyzed from multiple sources. Further, this study helps the researchers and investors to make a collective decision and choose the appropriate model for better profit and investment based on local and global market conditions.

Profit prediction optimization using financial accounting information system by optimized DLSTM

Financial accounting information systems (FAISs) are one of the scientific fields where deep learning (DL) and swarm-based algorithms have recently seen increased use. Nevertheless, the application of these hybrid networks has become more challenging as a result of the heightened complexity imposed by extensive datasets. In order to tackle this issue, we present a new methodology that integrates the twin adjustable reinforced chimp optimization algorithm (TAR-CHOA) with deep long short-term memory (DLSTM) to forecast profits using FAISs. The main contribution of this research is the development of the TAR-CHOA algorithm, which improves the efficacy of profit prediction models. Moreover, due to the unavailability of an appropriate dataset for this particular problem, a newly formed dataset has been constructed by employing fifteen inputs based on the prior Chinese stock market Kaggle dataset. In this study, we have designed and assessed five DLSTM-based optimization algorithms, for forecasting financial accounting profit. The performance of various models has been evaluated and ranked for financial accounting profit prediction. According to our research, the best-performing DL-based model is DLSTM-TAR-CHOA. One constraint of our methodology is its dependence on historical financial accounting data, operating under the assumption that past patterns and relationships will persist in the future. Furthermore, it is important to note that the efficacy of our models may differ based on the distinct attributes and fluctuations observed in various financial markets. These identified limitations present potential avenues for future research to investigate alternative methodologies and broaden the extent of our findings.

LSTM based stock prediction using weighted and categorized financial news

A significant correlation between financial news with stock market trends has been explored extensively. However, very little research has been conducted for stock prediction models that utilize news categories, weighted according to their relevance with the target stock. In this paper, we show that prediction accuracy can be enhanced by incorporating weighted news categories simultaneously into the prediction model. We suggest utilizing news categories associated with the structural hierarchy of the stock market: that is, news categories for the market, sector, and stock-related news. In this context, Long Short-Term Memory (LSTM) based Weighted and Categorized News Stock prediction model (WCN-LSTM) is proposed. The model incorporates news categories with their learned weights simultaneously. To enhance the effectiveness, sophisticated features are integrated into WCN-LSTM. These include, hybrid input, lexicon-based sentiment analysis, and deep learning to impose sequential learning. Experiments have been performed for the case of the Pakistan Stock Exchange (PSX) using different sentiment dictionaries and time steps. Accuracy and F1-score are used to evaluate the prediction model. We have analyzed the WCN-LSTM results thoroughly and identified that WCN-LSTM performs better than the baseline model. Moreover, the sentiment lexicon HIV4 along with time steps 3 and 7, optimized the prediction accuracy. We have conducted statistical analysis to quantitatively assess our findings. A qualitative comparison of WCN-LSTM with existing prediction models is also presented to highlight its superiority and novelty over its counterparts.

Reinforcement Learning for Quantitative Trading

Quantitative trading (QT), which refers to the usage of mathematical models and data-driven techniques in analyzing the financial market, has been a popular topic in both academia and financial industry since 1970s. In the last decade, reinforcement learning (RL) has garnered significant interest in many domains such as robotics and video games, owing to its outstanding ability on solving complex sequential decision making problems. RL’s impact is pervasive, recently demonstrating its ability to conquer many challenging QT tasks. It is a flourishing research direction to explore RL techniques’ potential on QT tasks. This paper aims at providing a comprehensive survey of research efforts on RL-based methods for QT tasks. More concretely, we devise a taxonomy of RL-based QT models, along with a comprehensive summary of the state of the art. Finally, we discuss current challenges and propose future research directions in this exciting field.

Leveraging machine learning and blockchain in E-commerce and beyond: benefits, models, and application

Blockchain technology (BT) allows market participants to keep track of digital transactions without central recordkeeping. The features of blockchain, including decentralization, persistency, and attack resistance, allow data security and privacy. Machine learning (ML) involves the analytical platform on a massive amount of data to provide precise decisions. Since data reliability, integration, and data security are crucial in machine learning, the emergence of blockchain technology and machine learning has become a unique, most disruptive, and trending research in the last few years, achieving comparable and precise performance. The combination of blockchain and machine learning (BT–ML) has been applied across different applications to assist decision-makers in retrieving valuable data insights while preserving privacy and integration. This paper summarizes the state-of-the-art research in combing BT and ML in e-commerce and other various applications, including healthcare, smart transportation, and the Internet of Things (IoT). The challenges and benefits of integrating machine learning and blockchain technologies are outlined in the paper. We also discuss the advantages and limitations of current algorithms in the BT–ML integration. This paper provides a roadmap for researchers to pave the way for current and future research directions in combing the BT and ML research areas.

Deep learning in the stock market-a systematic survey of practice, backtesting, and applications

The widespread usage of machine learning in different mainstream contexts has made deep learning the technique of choice in various domains, including finance. This systematic survey explores various scenarios employing deep learning in financial markets, especially the stock market. A key requirement for our methodology is its focus on research papers involving backtesting. That is, we consider whether the experimentation mode is sufficient for market practitioners to consider the work in a real-world use case. Works meeting this requirement are distributed across seven distinct specializations. Most studies focus on trade strategy, price prediction, and portfolio management, with a limited number considering market simulation, stock selection, hedging strategy, and risk management. We also recognize that domain-specific metrics such as "returns" and "volatility" appear most important for accurately representing model performance across specializations. Our study demonstrates that, although there have been some improvements in reproducibility, substantial work remains to be done regarding model explainability. Accordingly, we suggest several future directions, such as improving trust by creating reproducible, explainable, and accountable models and emphasizing prediction of longer-term horizons-potentially via the utilization of supplementary data-which continues to represent a significant unresolved challenge.

Systematic review of content analysis algorithms based on deep neural networks

Today according to social media, the internet, Etc. Data is rapidly produced and occupies a large space in systems that have resulted in enormous data warehouses; the progress in information technology has significantly increased the speed and ease of data flow.text mining is one of the most important methods for extracting a useful model through extracting and adapting knowledge from data sets. However, many studies have been conducted based on the usage of deep learning for text processing and text mining issues.The idea and method of text mining are one of the fields that seek to extract useful information from unstructured textual data that is used very today. Deep learning and machine learning techniques in classification and text mining and their type are discussed in this paper as well. Neural networks of various kinds, namely, ANN, RNN, CNN, and LSTM, are the subject of study to select the best technique. In this study, we conducted a Systematic Literature Review to extract and associate the algorithms and features that have been used in this area. Based on our search criteria, we retrieved 130 relevant studies from electronic databases between 1997 and 2021; we have selected 43 studies for further analysis using inclusion and exclusion criteria in Section 3.2. According to this study, hybrid LSTM is the most widely used deep learning algorithm in these studies, and SVM in machine learning method high accuracy in result shown.

Empirical mode decomposition using deep learning model for financial market forecasting

Financial market forecasting is an essential component of financial systems; however, predicting financial market trends is a challenging job due to noisy and non-stationary information. Deep learning is renowned for bringing out excellent abstract features from the huge volume of raw data without depending on prior knowledge, which is potentially fascinating in forecasting financial transactions. This article aims to propose a deep learning model that autonomously mines the statistical rules of data and guides the financial market transactions based on empirical mode decomposition (EMD) with back-propagation neural networks (BPNN). Through the characteristic time scale of data, the intrinsic wave pattern was obtained and then decomposed. Financial market transaction data were analyzed, optimized using PSO, and predicted. Combining the nonlinear and non-stationary financial time series can improve prediction accuracy. The predictive model of deep learning, based on the analysis of the massive financial trading data, can forecast the future trend of financial market price, forming a trading signal when particular confidence is satisfied. The empirical results show that the EMD-based deep learning model has an excellent predicting performance.

Artificial intelligence in multi-parametric magnetic resonance imaging: A review

Multi-parametric magnetic resonance imaging (mpMRI) is an indispensable tool in the clinical workflow for the diagnosis and treatment planning of various diseases. Machine learning-based artificial intelligence (AI) methods, especially those adopting the deep learning technique, have been extensively employed to perform mpMRI image classification, segmentation, registration, detection, reconstruction, and super-resolution. The current availability of increasing computational power and fast-improving AI algorithms have empowered numerous computer-based systems for applying mpMRI to disease diagnosis, imaging-guided radiotherapy, patient risk and overall survival time prediction, and the development of advanced quantitative imaging technology for magnetic resonance fingerprinting. However, the wide application of these developed systems in the clinic is still limited by a number of factors, including robustness, reliability, and interpretability. This survey aims to provide an overview for new researchers in the field as well as radiologists with the hope that they can understand the general concepts, main application scenarios, and remaining challenges of AI in mpMRI. This article is protected by copyright. All rights reserved.

Real Estate Tax Base Assessment by Deep Learning Neural Network in the Context of the Digital Economy

With the continuous development of China's digital economy and the continuous heating of the real estate market, real estate tax base assessment occupies an important position in the real estate market. The purpose is to improve the work efficiency of relevant personnel of real estate tax base assessment, reduce workload pressure, and improve the evaluation level. Real estate tax base assessment and real estate appraisal are studied in detail, and the factors of the real estate tax base assessment index are analyzed. Different real estate tax base assessment methods are compared, and the difference and connection between different methods are explored. The theory of batch assessment of real estate tax base is analyzed in depth, and the procedures for batch assessment implementation are summarized. On this basis, a deep learning neural network (DLNN) theory is proposed, and a real estate tax base assessment model based on DLNN is constructed. The reliability, accuracy, and relative superiority of the model are analyzed in detail, and the model is used to test the sample data and analyze the error. The results reveal that the DLNN model has better data fit and good reliability. Compared with other algorithms, it has certain advantages and smaller error values. In the sample test, the test value is closer to the actual value, the error is controllable, and it has high accuracy. Through training, it shows that the DL model has an excellent performance in tax base assessment, can meet the requirements of efficient batch assessment, and is expected to achieve the goal of completing a huge workload in a limited time and improve work efficiency. The real estate tax base assessment model by DLNN can bring some help to the real estate finance and taxation work and provide a reference for the batch assessment of tax base in the real estate industry.

Incorporating causality in energy consumption forecasting using deep neural networks

Forecasting energy demand has been a critical process in various decision support systems regarding consumption planning, distribution strategies, and energy policies. Traditionally, forecasting energy consumption or demand methods included trend analyses, regression, and auto-regression. With advancements in machine learning methods, algorithms such as support vector machines, artificial neural networks, and random forests became prevalent. In recent times, with an unprecedented improvement in computing capabilities, deep learning algorithms are increasingly used to forecast energy consumption/demand. In this contribution, a relatively novel approach is employed to use long-term memory. Weather data was used to forecast the energy consumption from three datasets, with an additional piece of information in the deep learning architecture. This additional information carries the causal relationships between the weather indicators and energy consumption. This architecture with the causal information is termed as entangled long short term memory. The results show that the entangled long short term memory outperforms the state-of-the-art deep learning architecture (bidirectional long short term memory). The theoretical and practical implications of these results are discussed in terms of decision-making and energy management systems.

Reconstructing the Local Volatility Surface from Market Option Prices

We present an efficient and accurate computational algorithm for reconstructing a local volatility surface from given market option prices. The local volatility surface is dependent on the values of both the time and underlying asset. We use the generalized Black–Scholes (BS) equation and finite difference method (FDM) to numerically solve the generalized BS equation. We reconstruct the local volatility function, which provides the best fit between the theoretical and market option prices by minimizing a cost function that is a quadratic representation of the difference between the two option prices. This is an inverse problem in which we want to calculate a local volatility function consistent with the observed market prices. To achieve robust computation, we place the sample points of the unknown volatility function in the middle of the expiration dates. We perform various numerical experiments to confirm the simplicity, robustness, and accuracy of the proposed method in reconstructing the local volatility function.

A Review of Sentiment, Semantic and Event-Extraction-Based Approaches in Stock Forecasting

Stock forecasting is a significant and challenging task. The recent development of web technologies has transformed the communication channel to allow the public to share information over the web such as news, social media contents, etc., thus causing exponential growth of web data. The massively available information might be the key to revealing the financial market’s unexplained variability and facilitating forecasting accuracy. However, this information is usually in unstructured natural language and consists of different inherent meanings. Although a human can easily interpret the inherent messages, it is still complicated to manually process such a massive amount of textual data due to the constraint of time, ability, energy, etc. Due to the different properties of text sources, it is crucial to understand various text processing approaches to optimize forecasting performance. This study attempted to summarize and discuss the current text-based financial forecasting approaches in the aspect of semantic-based, sentiment-based, event-extraction-based, and hybrid approaches. Afterward, the study discussed the strength and weakness of each approach, followed with their comparison and suitable application scenarios. Moreover, this study also highlighted the future research direction in text-based stock forecasting, where the overall discussion is expected to provide insightful analysis for future reference.

Deep learning model based on urban multi-source data for predicting heavy metals (Cu, Zn, Ni, Cr) in industrial sewer networks

The high concentrations of heavy metals in municipal industrial sewer networks will seriously impact the microorganisms of the activated sludge in the wastewater treatment plant (WWTP), thus deteriorating the effluent quality and destroying the stability of sewage treatment. Therefore, timely prediction and early warning of heavy metal concentrations in industrial sewer networks is crucial. However, due to the complex sources of heavy metals in industrial sewer networks, traditional physical modeling and linear methods cannot establish an accurate prediction model. Herein, we developed a Gated Recurrent Unit (GRU) neural network model based on a deep learning algorithm for predicting the concentrations of heavy metals in industrial sewer networks. To train the GRU model, we used low-cost and easy-to-obtain urban multi-source data, including socio-environmental indicator data, air environmental indicator data, water quantity indicator data, and easily measurable water quality indicator data. The model was applied to predict the concentrations of heavy metals (Cu, Zn, Ni, and Cr) in the sewer networks of an industrial area in southern China. The results are compared with the commonly used Artificial Neural Network (ANN) model. In this study, it was shown that the GRU had better prediction performance for Cu, Zn, Ni, and Cr concentrations, with the average R² significantly increased by 12.35%, 11.94%, 9.21%, and 8.13%, respectively, compared to ANN predictions. The sensitivity analysis based on Shapley (SHAP) values revealed that conductivity (σ), temperature (T), pH, and sewage flow (Flow) contributed significantly to the prediction results of the model. Furthermore, the three input variables including air pressure (AP), land area (A), and population (Pop.) were removed without affecting the prediction performance of the model, which maximized the modeling efficiency and reduced the operational cost. This study provides an economical and feasible technical method for early warning of abnormal heavy metal concentrations in urban industrial sewer networks.

A Comprehensive Study of Market Prediction from Efficient Market Hypothesis up to Late Intelligent Market Prediction Approaches

This paper has scrutinized the process of testing market efficiency, data generation process and the feasibility of market prediction with a detailed, coherent and statistical approach. Furthermore, attempts are made to extract knowledge from S&P 500 market data with an emphasize on feature engineering. As such, different data representations are provided through different procedures, and their performance in knowledge extraction is discussed. Amongst the neural networks, Long Short-Term Memory has not been adequately experimented. LSTM, because of its intrinsic, considers the long-term and short-term memory in its computations. Thus, in this paper LSTM is further examined in return prediction and different preprocessing methods are tested to improve its accuracy. This study is conducted on market data during September-2000 to February-2021. In order to extend the amount of knowledge extracted from financial time series, and to select the best input features, the advantage of Principal Component Analyze, Random Forest, Wavelet and the LSTM's own deep feature extraction procedure are taken, and 4 models are compiled. Subsequently, to validate the performance of the models, MAE, MSE, MAPE, CSP and CDCP are calculated. Results from Diebold Mariano test implied that although LSTM neural network has gained a lot of attention recently, it does not significantly perform better than the benchmark method in S&P 500 index return prediction. Yet, results from Wilcoxon signed rank test showed the significance of improvement in the predictions performed by the combination of Principal component analysis and LSTM.

Clustering-enhanced stock price prediction using deep learning

In recent years, artificial intelligence technologies have been successfully applied in time series prediction and analytic tasks. At the same time, a lot of attention has been paid to financial time series prediction, which targets the development of novel deep learning models or optimize the forecasting results. To optimize the accuracy of stock price prediction, in this paper, we propose a clustering-enhanced deep learning framework to predict stock prices with three matured deep learning forecasting models, such as Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN) and Gated Recurrent Unit (GRU). The proposed framework considers the clustering as the forecasting pre-processing, which can improve the quality of the training models. To achieve the effective clustering, we propose a new similarity measure, called Logistic Weighted Dynamic Time Warping (LWDTW), by extending a Weighted Dynamic Time Warping (WDTW) method to capture the relative importance of return observations when calculating distance matrices. Especially, based on the empirical distributions of stock returns, the cost weight function of WDTW is modified with logistic probability density distribution function. In addition, we further implement the clustering-based forecasting framework with the above three deep learning models. Finally, extensive experiments on daily US stock price data sets show that our framework has achieved excellent forecasting performance with overall best results for the combination of Logistic WDTW clustering and LSTM model using 5 different evaluation metrics.

Electric Shovel Teeth Missing Detection Method Based on Deep Learning

Electric shovels are widely used in the mining industry to dig ore, and the teeth in shovels' bucket can be lost due to the tremendous pressure exerted by ore materials during operation. When the teeth fall off and enter the crusher with other ore materials, serious damages to crusher gears and other equipment happen, which causes millions of economic loss, because it is made of high-manganese steel. Thus, it is urgent to develop an efficient and automatic algorithm for detecting broken teeth. However, existing methods for detecting broken teeth have little effect and most research studies depended on sensor skills, which will be disturbed by closed cavity in shovel and not stable in practice. In this paper, we present an intelligent computer vision system for monitoring teeth condition and detecting missing teeth. Since the pixel-level algorithm is carried out, the amount of calculation should be reduced to improve the superiority of the algorithm. To release computational pressure of subsequent work, salient detection based on deep learning is proposed for extracting the key frame images from video flow taken by the camera installed on the shovel including the teeth we intend to analyze. Additionally, in order to more efficiently monitor teeth condition and detect missing teeth, semantic segmentation based on deep learning is processed to get the relative position of the teeth in the image. Once semantic segmentation is done, floating images containing the shape of teeth are obtained. Then, to detect missing teeth effectively, image registration is proposed. Finally, the result of image registration shows whether teeth are missing or not, and the system will immediately alert staff to check the shovel when teeth fall off. Through sufficient experiments, statistical result had demonstrated superiority of our presented model that serves more promising prospect in mining industry.

Adaptive Localizing Region-Based Level Set for Segmentation of Maxillary Sinus Based on Convolutional Neural Networks

In this paper, we propose a novel method, an adaptive localizing region-based level set using convolutional neural network, for improving performance of maxillary sinus segmentation. The healthy sinus without lesion inside is easy for conventional algorithms. However, in practice, most of the cases are filled with lesions of great heterogeneity which lead to lower accuracy. Therefore, we provide a strategy to avoid active contour from being trapped into a nontarget area. First, features of lesion and maxillary sinus are studied using a convolutional neural network (CNN) with two convolutional and three fully connected layers in architecture. In addition, outputs of CNN are devised to evaluate possibilities of zero level set location close to lesion or not. Finally, the method estimates stable points on the contour by an interactive process. If it locates in the lesion, the point needs to be paid a certain speed compensation based on the value of possibility via CNN, assisting itself to escape from the local minima. If not, the point preserves current status till convergence. Capabilities of our method have been demonstrated on a dataset of 200 CT images with possible lesions. To illustrate the strength of our method, we evaluated it against state-of-the-art methods, FLS and CRF-FCN. For all cases, our method, as assessed by Dice similarity coefficients, performed significantly better compared with currently available methods and obtained a significant Dice improvement, 0.25 than FLS and 0.12 than CRF-FCN, respectively, on an average.

COVID-19 outbreak: An ensemble pre-trained deep learning model for detecting informative tweets

On 11 March 2020, the (WHO) World Health Organization declared COVID-19 (CoronaVirus Disease 2019) as a pandemic. A further crisis has manifested mass fear and panic, driven by lack of information, or sometimes outright misinformation, alongside the coronavirus pandemic. Twitter is one of the prominent and trusted social media in this current outbreak. Over time, boundless COVID-19 headlines and vast awareness have been spreading, with tweets, updates, videos, and explosive posts. Few studies have been performed on the pandemic to detect and interrelate various disease types, including current coronavirus. However, it is pretty tricky to discriminate and detect a specific category. This work is motivated by the need to inform society about limiting irrelevant information and avoiding spreading negative emotions. In this context, the current work focuses on informative tweet detection in the pandemic to provide relevant information to the government, medical organizations, victims services, etc. This paper used a Majority Voting technique-based Ensemble Deep Learning (MVEDL) model. This MVEDL model is used to identify COVID-19 related (INFORMATIVE) tweets. The state-of-art deep learning models RoBERTa, BERTweet, and CT-BERT are used for best performance with the MVEDL model. The "COVID-19 English labeled tweets" dataset is used for training and testing the MVEDL model. The MVEDL model has shown 91.75 percent accuracy, 91.14 percent F1-score and outperforms the traditional machine learning and deep learning models. We also investigate how to use the MVEDL model for sentiment analysis on 226668 unlabeled COVID-19 tweets and their informative tweets. The application section discussed a comprehensive analysis of both actual and informative tweets. According to our knowledge, this is the first work on COVID-19 sentiment analysis using a deep learning ensemble model.

DPP: Deep predictor for price movement from candlestick charts

Forecasting the stock market prices is complicated and challenging since the price movement is affected by many factors such as releasing market news about earnings and profits, international and domestic economic situation, political events, monetary policy, major abrupt affairs, etc. In this work, a novel framework: deep predictor for price movement (DPP) using candlestick charts in the stock historical data is proposed. This framework comprises three steps: 1. decomposing a given candlestick chart into sub-charts; 2. using CNN-autoencoder to acquire the best representation of sub-charts; 3. applying RNN to predict the price movements from a collection of sub-chart representations. An extensive study is operated to assess the performance of the DPP based models using the trading data of Taiwan Stock Exchange Capitalization Weighted Stock Index and a stock market index, Nikkei 225, for the Tokyo Stock Exchange. Three baseline models based on IEM, Prophet, and LSTM approaches are compared with the DPP based models.

Stock Price Movement Prediction Based on a Deep Factorization Machine and the Attention Mechanism

The prediction of stock price movement is a popular area of research in academic and industrial fields due to the dynamic, highly sensitive, nonlinear and chaotic nature of stock prices. In this paper, we constructed a convolutional neural network model based on a deep factorization machine and attention mechanism (FA-CNN) to improve the prediction accuracy of stock price movement via enhanced feature learning. Unlike most previous studies, which focus only on the temporal features of financial time series data, our model also extracts intraday interactions among input features. Further, in data representation, we used the sub-industry index as supplementary information for the current state of the stock, since there exists stock price co-movement between individual stocks and their industry index. The experiments were carried on the individual stocks in three industries. The results showed that the additional inputs of (a) the intraday interactions among input features and (b) the sub-industry index information effectively improved the prediction accuracy. The highest prediction accuracy of the proposed FA-CNN model is 64.81%. It is 7.38% higher than that of traditional LSTM, and 3.71% higher than that of the model without sub-industry index as additional input features.

TNT: An Interpretable Tree-Network-Tree Learning Framework using Knowledge Distillation

Deep Neural Networks (DNNs) usually work in an end-to-end manner. This makes the trained DNNs easy to use, but they remain an ambiguous decision process for every test case. Unfortunately, the interpretability of decisions is crucial in some scenarios, such as medical or financial data mining and decision-making. In this paper, we propose a Tree-Network-Tree (TNT) learning framework for explainable decision-making, where the knowledge is alternately transferred between the tree model and DNNs. Specifically, the proposed TNT learning framework exerts the advantages of different models at different stages: (1) a novel James-Stein Decision Tree (JSDT) is proposed to generate better knowledge representations for DNNs, especially when the input data are in low-frequency or low-quality; (2) the DNNs output high-performing prediction result from the knowledge embedding inputs and behave as a teacher model for the following tree model; and (3) a novel distillable Gradient Boosted Decision Tree (dGBDT) is proposed to learn interpretable trees from the soft labels and make a comparable prediction as DNNs do. Extensive experiments on various machine learning tasks demonstrated the effectiveness of the proposed method.