Cardiac Arrhythmia Classification Using Advanced Deep Learning Techniques on Digitized ECG Datasets

Sensors, Journal Year: 2024, Volume and Issue: 24(8), P. 2484 - 2484

Published: April 12, 2024

ECG classification or heartbeat is an extremely valuable tool in cardiology. Deep learning-based techniques for the analysis of signals assist human experts timely diagnosis cardiac diseases and help save precious lives. This research aims at digitizing a dataset images records into time series then applying deep learning (DL) on digitized dataset. State-of-the-art DL are proposed different classes. Multiple models, including convolutional neural network (CNN), long short-term memory (LSTM) network, self-supervised (SSL)-based model using autoencoders explored compared this study. The models trained generated from plots patients various healthcare institutes Pakistan. First, digitized, segmenting lead II heartbeats, passed to classification. Among used study, CNN achieves highest accuracy ∼92%. highly accurate provides fast inference real-time direct monitoring that captured electrodes (sensors) placed parts body. Using form instead arrhythmia allows cardiologists utilize directly machine ECGs.

Language: Английский

---

Framework for Predicting and Modeling Stock Market Prices Based on Deep Learning Algorithms

Electronics, Journal Year: 2022, Volume and Issue: 11(19), P. 3149 - 3149

Published: Sept. 30, 2022

The creation of trustworthy models the equities market enables investors to make better-informed choices. A trading model may lessen risks that are connected with investing and it possible for traders choose companies offer highest dividends. However, due high degree correlation between stock prices, analysis is made more difficult by batch processing approaches. prediction has entered a technologically advanced era advent technological marvels such as global digitization. For this reason, artificial intelligence have become very important continuous increase in capitalization. novelty proposed study development robustness time series based on deep leaning forecasting future values marketing. primary purpose was develop an intelligent framework capability predicting direction which prices will move financial inputs. Among cutting-edge technologies, backbone many different predict markets. In particular, learning strategies been effective at behavior. article, we propose long short-term memory (LSTM) hybrid convolutional neural network (CNN-LSTM) LSTM closing Tesla, Inc. Apple, These predictions were using data collected over past two years. mean squared error (MSE), root (RMSE), normalization (NRMSE), Pearson’s (R) measures used computation findings models. Between models, CNN-LSTM scored slightly better (Tesla: R-squared = 98.37%; Apple: 99.48%). showed superior performance compared single existing systems prices.

Language: Английский

---

A Novel Deep Learning Technique for Detecting Emotional Impact in Online Education

Electronics, Journal Year: 2022, Volume and Issue: 11(18), P. 2964 - 2964

Published: Sept. 19, 2022

Emotional intelligence is the automatic detection of human emotions using various intelligent methods. Several studies have been conducted on emotional intelligence, and only a few adopted in education. Detecting student can significantly increase productivity improve education process. This paper proposes new deep learning method to detect emotions. The main aim this map relationship between teaching practices based impact. Facial recognition algorithms extract helpful information from online platforms as image classification techniques are applied and/or teacher faces. As part work, two models compared according their performance. Promising results achieved both techniques, presented Experimental Results Section. For validation proposed system, an course with students used; findings suggest that technique operates well. Based analysis, several train test emotion Transfer for pre-trained neural network used well accuracy stage. obtained show performance promising

Language: Английский

---

Optimization of Traditional Stock Market Strategies Using the LSTM Hybrid Approach

Information, Journal Year: 2024, Volume and Issue: 15(3), P. 136 - 136

Published: Feb. 28, 2024

Investment decision-makers increasingly rely on modern digital technologies to enhance their strategies in today’s rapidly changing and complex market environment. This paper examines the impact of incorporating Long Short-term Memory (LSTM) models into traditional trading strategies. The core investigation revolves around whether enhanced with LSTM technology perform better than methods alone. Traditional typically depend analyzing current closing prices various technical indicators take action. However, by applying models, this study aims forecast greater accuracy, thereby improving performance. Our findings indicate that utilize outperform improvement suggests a significant advantage using for prediction decision making. Acknowledging no one-size-fits-all strategy works every condition or stock is crucial. As such, traders are encouraged select tailor based thorough testing analysis best suit needs conditions. contributes understanding how integrating can strategies, offering path toward more effective making unpredictable market.

Language: Английский

---

Forecasting stock prices of fintech companies of India using random forest with high-frequency data

Journal of Open Innovation Technology Market and Complexity, Journal Year: 2023, Volume and Issue: 10(1), P. 100180 - 100180

Published: Nov. 18, 2023

The fintech segment is currently one of the most rapidly growing industries, attracting numerous investors who anticipate substantial returns in future. Notably, not only individual retail but also mutual fund agencies are actively engaged predicting stock prices within this sector to maximize their trading gains. purpose study formulate forecasting models for top three Fintech Companies India i.e., Policy Bazar, One 97 Communications Paytm Ltd., and Niyogin Ltd. Using Random Forest model with high-frequency data Python. literature review section proves that a novel piece work as none existing research focused on using model. extracted from www.moneycontrol.com www.kotaksecurities.com, period 1st October, 2022 30th September, 2023. deals about 293,280 points 3 companies @ 97,760 each. It has been found random forest provides very successful results prediction co-efficient determination all selected more than 95%.

Language: Английский

---

Forecasting multistep daily stock prices for long-term investment decisions: A study of deep learning models on global indices

Engineering Applications of Artificial Intelligence, Journal Year: 2023, Volume and Issue: 129, P. 107617 - 107617

Published: Dec. 3, 2023

Language: Английский

---

Deep-SDM: A Unified Computational Framework for Sequential Data Modeling Using Deep Learning Models

Software, Journal Year: 2024, Volume and Issue: 3(1), P. 47 - 61

Published: Feb. 28, 2024

Deep-SDM is a unified layer framework built on TensorFlow/Keras and written in Python 3.12. The aligns with the modular engineering principles for design development strategy. Transparency, reproducibility, recombinability are framework’s primary criteria. platform can extract valuable insights from numerical text data utilize them to predict future values by implementing long short-term memory (LSTM), gated recurrent unit (GRU), convolution neural network (CNN). Its end-to-end machine learning pipeline involves sequence of tasks, including exploration, input preparation, model construction, hyperparameter tuning, performance evaluations, visualization results, statistical analysis. complete process systematic carefully organized, import selection, encapsulating it into whole. multiple subroutines work together provide user-friendly conducive that easy use. We utilized Nepal Stock Exchange (NEPSE) index validate its reproducibility robustness observed impressive results.

Language: Английский

---

Long Short-Term Memory and Gated Recurrent Unit for Stock Price Prediction

Procedia Computer Science, Journal Year: 2024, Volume and Issue: 234, P. 204 - 212

Published: Jan. 1, 2024

Stocks are a popular investment with high risk due to rapid price fluctuations that difficult predict. Many investors do not understand the analysis of buying and selling stocks, making them hesitant invest. For this reason, an analytical technique is needed can determine movement stock prices in order carry out planning, management, decision-making. Banking stocks among important sectors. One go public banking Bank Rakyat Indonesia stock. This research applies Long Short-Term Memory Gated Recurrent Unit produce model accurately predict Indonesia. Based on implementation, GRU best MSE value 4958.9168, RMSE 70.4195, MAPE 1.1699%. The there will be decrease next month.

Language: Английский

---

Harnessing Deep Learning and Technical Indicators for Enhanced Stock Predictions of Blue-Chip Stocks on the Indonesia Stock Exchange (IDX)

Engineering Technology & Applied Science Research, Journal Year: 2025, Volume and Issue: 15(1), P. 20348 - 20357

Published: Feb. 2, 2025

Given the limitations of existing models in accurately predicting stock prices, particularly emerging markets such as Indonesia, this study aimed to evaluate effectiveness deep learning forecasting prices using blue-chip company shares traded on Indonesia Stock Exchange (IDX). The main focus lies combining historical data with a series technical indicators, optimizing their integration improve prediction accuracy. accuracy method is reflected comprehensive evaluation model performance robust metrics, including R2, Mean Squared Error (MSE), and Root (RMSE). Empirical results show superiority integrating indicators compared relying only data. LSTM showed most significant improvement, R2 for ASII jumping by 14.59% after incorporating indicators. GRU BBCA increased significantly, shown decrease 45.16% MSE. These findings underscore critical role feature selection developing models. Integrating increases provides additional tools informed decision-making.

Language: Английский

---

Forecasting Covered Call Exchange-Traded Funds (ETFs) Using Time Series, Machine Learning, and Deep Learning Models

Journal of risk and financial management, Journal Year: 2025, Volume and Issue: 18(3), P. 120 - 120

Published: Feb. 25, 2025

This study explores the application of time series, machine learning (ML), and deep (DL) models to predict prices performance covered call ETFs. Utilizing historical data from major ETFs like QYLD, XYLD, JEPI, JEPQ, RYLD, research assesses predictive accuracy reliability different forecasting approaches. It compares traditional series methods, including ARIMA Heterogeneous Autoregressive Model (HAR), with advanced ML techniques such as Random Forests (RF) Support Vector Regression (SVR), well DL Recurrent Neural Networks (RNN) Convolutional (CNN). is evaluated using metrics Mean Absolute Error (MAE), Root Square (RMSE), Percentage (MAPE), Akaike Information Criterion (AIC), Bayesian (BIC). Results indicate that are effective at identifying nonlinear patterns temporal dependencies in price movements ETFs, outperforming both techniques. These findings enhance existing financial literature offer valuable insights for investors portfolio managers aiming improve their strategies

Language: Английский

---