Deep Learning for Credit Card Fraud Detection: A Review of Algorithms, Challenges, and Solutions

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 96893 - 96910

Published: Jan. 1, 2024

Deep learning (DL), a branch of machine (ML), is the core technology in today's technological advancements and innovations. learning-based approaches are state-of-the-art methods used to analyse detect complex patterns large datasets, such as credit card transactions. However, most fraud models literature based on traditional ML algorithms, recently, there has been rise applications deep techniques. This study reviews recent DL-based presents concise description performance comparison widely DL techniques, including convolutional neural network (CNN), simple recurrent (RNN), long short-term memory (LSTM), gated unit (GRU). Additionally, an attempt made discuss suitable metrics, common challenges encountered when training using architectures potential solutions, which lacking previous studies would benefit researchers practitioners. Meanwhile, experimental results analysis real-world dataset indicate robustness detection.

Language: Английский

---

Accurate Stock Price Forecasting Based on Deep Learning and Hierarchical Frequency Decomposition

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 49878 - 49894

Published: Jan. 1, 2024

The stock market plays an increasingly important role in the global economy. Accurate price forecasting not only aids government predicting economic trends, but also helps investors anticipate higher expected returns. Nevertheless, hurdles such as non-linearity, complexity and high volatility make it a daunting task to predict prices. To address this issue, paper proposes new hybrid model, termed Hierarchical Decomposition based Forecasting Model (HDFM), decompose forecast prices hierarchical fashion. model utilises complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) for initial of time series. enhance predictive efficiency, sub-series similar sample entropy from are combined K-means clustering method. Through thorough analysis, is found that first contains more high-frequency signals. Therefore, subjected second variational (VMD). Afterwards, gated recurrent unit (GRU) used each individually. final results obtained by merging prediction outcomes. proposed has been evaluated on three different markets. experimental showed outperformed other methods across all indices. Moreover, ablation studies demonstrated effectiveness individual component within model.

Language: Английский

---

Multi-factor stock trading strategy based on DQN with multi-BiGRU and multi-head ProbSparse self-attention

Applied Intelligence, Journal Year: 2024, Volume and Issue: 54(7), P. 5417 - 5440

Published: April 1, 2024

Abstract Reinforcement learning is widely used in financial markets to assist investors developing trading strategies. However, most existing models primarily focus on simple volume-price factors, and there a need for further improvement the returns of stock trading. To address these challenges, multi-factor strategy based Deep Q-Network (DQN) with Multi-layer Bidirectional Gated Recurrent Unit (Multi-BiGRU) multi-head ProbSparse self-attention proposed. Our comprehensively characterizes determinants prices by considering various factors such as quality, valuation, sentiment factors. We first use Light Gradient Boosting Machine (LightGBM) classify turning points data. Then, reinforcement strategy, Multi-BiGRU, which holds bidirectional historical data, integrated into DQN, aiming enhance model’s ability understand dynamics market. Moreover, mechanism effectively captures interactions between different providing model deeper market insights. validate our strategy’s effectiveness through extensive experimental research stocks from Chinese US markets. The results show that method outperforms both temporal non-temporal terms returns. Ablation studies confirm critical role LightGBM mechanism. experiment section also demonstrates significant advantages presentation box plots statistical tests. Overall, fully data feature extraction capabilities, work expected provide more precise decision support. Graphical abstract

Language: Английский

---

A Multi-Feature Selection Fused with Investor Sentiment for Stock Price Prediction

Published: Jan. 1, 2025

Language: Английский

---

Artificial intelligence-driven financial innovation: A robo-advisor system for robust returns across diversified markets

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 126881 - 126881

Published: Feb. 1, 2025

Language: Английский

---

A multi-feature selection fused with investor sentiment for stock price prediction

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127381 - 127381

Published: March 1, 2025

Language: Английский

---

GHENet: Attention-based Hurst exponents for the forecasting of stock market indexes

Physica A Statistical Mechanics and its Applications, Journal Year: 2025, Volume and Issue: unknown, P. 130540 - 130540

Published: March 1, 2025

Language: Английский

---

Piyasa Yönünün Derin Öğrenme ile Tahmini: E-7 Ülke Borsaları Üzerine Bir Uygulama

Maliye Finans Yazıları, Journal Year: 2025, Volume and Issue: 123, P. 92 - 111

Published: April 1, 2025

Bu çalışmada, derin öğrenme yönteminin yükselen piyasa ekonomileri olarak bilinen E-7 ülkelerinin borsa endeksleri üzerindeki tahmin performansının incelenmesi amaçlanmıştır. bağlamda, IPC (Meksika), SSE (Çin), BIST 100 (Türkiye), RTS (Rusya), BOVESPA (Brezilya), IDX (Endonezya) ve NIFTY 50 (Hindistan) endekslerinin günlük hareket yönleri H2O modeli kullanılarak edilmiştir. Modelin girdileri olarak, 01.01.2015 31.12.2024 tarihleri arasında kapanış fiyatlarına dayalı hesaplanan teknik göstergeler kullanılmıştır. Tahmin sürecinde veriler %80 eğitim %20 test seti bölünmüştür. Hesaplanan doğruluk oranları endeksi için %88,47, %78,13, %77,29, %76,05, %75,81, %75,05 %74,34 bulunmuştur. Elde edilen bulgular, yöntemlerinin hareketlerini belirli bir düzeyiyle edebildiğini göstermektedir.

---

A novel decision ensemble framework: Attention-customized BiLSTM and XGBoost for speculative stock price forecasting

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(4), P. e0320089 - e0320089

Published: April 16, 2025

Forecasting speculative stock prices is essential for effective investment risk management and requires innovative algorithms. However, the nature, volatility, complex sequential dependencies within financial markets present inherent challenges that necessitate advanced techniques. In this regard, a novel framework, ACB-XDE (Attention-Customized BiLSTM-XGB Decision Ensemble), proposed predicting daily closing price of Bitcoin-USD (BTC-USD). The framework integrates learning capabilities customized Bi-directional Long Short-Term Memory (BiLSTM) model with attention mechanism XGBoost algorithm. BiLSTM leverages its to capture market trends. Meanwhile, new dynamically assigns weights influential features based on volatility patterns, thereby enhancing interpretability optimizing cost measures forecasting. Moreover, handles nonlinear relationships contributes framework’s robustness. Furthermore, error reciprocal method improves predictions by iteratively adjusting difference between theoretical expectations actual errors in individual attention-customized models. Finally, from both models are concatenated create varied prediction space, which then fed into ensemble regression improve generalization framework. Empirical validation involves application volatile Bitcoin market, utilizing dataset sourced Yahoo Finance (Bitcoin-USD, 10/01/2014 01/08/2023). outperforms state-of-the-art MAPE 0.37%, MAE 84.40, RMSE 106.14. This represents improvements approximately 27.45%, 53.32%, 38.59% MAPE, MAE, respectively, over best-performing attention-BiLSTM. presents technique informed decision-making dynamic landscapes demonstrates effectiveness handling complexities BTC-USD data.

Language: Английский

---

Time Series Prediction for Cryptocurrency Markets with Transformer and Parallel Convolutional Neural Networks

Applied Soft Computing, Journal Year: 2025, Volume and Issue: unknown, P. 113229 - 113229

Published: May 1, 2025

Language: Английский

---