A review on the long short-term memory model

Artificial Intelligence Review, Journal Year: 2020, Volume and Issue: 53(8), P. 5929 - 5955

Published: May 13, 2020

Language: Английский

---

A comprehensive review on ensemble deep learning: Opportunities and challenges

Journal of King Saud University - Computer and Information Sciences, Journal Year: 2023, Volume and Issue: 35(2), P. 757 - 774

Published: Feb. 1, 2023

In machine learning, two approaches outperform traditional algorithms: ensemble learning and deep learning. The former refers to methods that integrate multiple base models in the same framework obtain a stronger model outperforms them. success of an method depends on several factors, including how baseline are trained they combined. literature, there common building successfully applied domains. On other hand, learning-based have improved predictive accuracy across wide range Despite diversity architectures their ability deal with complex problems extract features automatically, main challenge is it requires lot expertise experience tune optimal hyper-parameters, which makes tedious time-consuming task. Numerous recent research efforts been made approach overcome this challenge. Most these focus simple some limitations. Hence, review paper provides comprehensive reviews various strategies for especially case Also, explains detail or factors influence methods. addition, presents accurately categorized used

Language: Английский

---

Deep learning and medical image processing for coronavirus (COVID-19) pandemic: A survey

Sustainable Cities and Society, Journal Year: 2020, Volume and Issue: 65, P. 102589 - 102589

Published: Nov. 5, 2020

Language: Английский

---

Feature dimensionality reduction: a review

Complex & Intelligent Systems, Journal Year: 2022, Volume and Issue: 8(3), P. 2663 - 2693

Published: Jan. 21, 2022

Abstract As basic research, it has also received increasing attention from people that the “curse of dimensionality” will lead to increase cost data storage and computing; influences efficiency accuracy dealing with problems. Feature dimensionality reduction as a key link in process pattern recognition become one hot difficulty spot field recognition, machine learning mining. It is most challenging research fields, which been favored by scholars’ attention. How implement “low loss” feature dimension reduction, keep nature original data, find out best mapping get optimal low dimensional are keys aims research. In this paper, two-dimensionality methods, selection extraction, introduced; current mainstream algorithms analyzed, including method for small sample based on deep learning. For each algorithm, examples their application given advantages disadvantages these methods evaluated.

Language: Английский

---

EEG emotion recognition using fusion model of graph convolutional neural networks and LSTM

Applied Soft Computing, Journal Year: 2020, Volume and Issue: 100, P. 106954 - 106954

Published: Dec. 1, 2020

Language: Английский

---

Deep learning modelling techniques: current progress, applications, advantages, and challenges

Artificial Intelligence Review, Journal Year: 2023, Volume and Issue: 56(11), P. 13521 - 13617

Published: April 17, 2023

Abstract Deep learning (DL) is revolutionizing evidence-based decision-making techniques that can be applied across various sectors. Specifically, it possesses the ability to utilize two or more levels of non-linear feature transformation given data via representation in order overcome limitations posed by large datasets. As a multidisciplinary field still its nascent phase, articles survey DL architectures encompassing full scope are rather limited. Thus, this paper comprehensively reviews state-of-art modelling and provides insights into their advantages challenges. It was found many models exhibit highly domain-specific efficiency could trained methods. However, training very time-consuming, expensive, requires huge samples for better accuracy. Since also susceptible deception misclassification tends get stuck on local minima, improved optimization parameters required create robust models. Regardless, has already been leading groundbreaking results healthcare, education, security, commercial, industrial, as well government Some models, like convolutional neural network (CNN), generative adversarial networks (GAN), recurrent (RNN), recursive networks, autoencoders, frequently used, while potential other remains widely unexplored. Pertinently, hybrid conventional have capacity challenges experienced Considering capsule may dominate future work aimed compile information stakeholders involved development use contemporary world.

Language: Английский

---

Financial Fraud: A Review of Anomaly Detection Techniques and Recent Advances

Expert Systems with Applications, Journal Year: 2021, Volume and Issue: 193, P. 116429 - 116429

Published: Dec. 31, 2021

With the rise of technology and continued economic growth evident in modern society, acts fraud have become much more prevalent financial industry, costing institutions consumers hundreds billions dollars annually. Fraudsters are continuously evolving their approaches to exploit vulnerabilities current prevention measures place, many whom targeting sector. These crimes include credit card fraud, healthcare automobile insurance money laundering, securities commodities insider trading. On own, systems do not provide adequate security against these criminal acts. As such, need for detection detect fraudulent after they already been committed potential cost savings doing so is than ever. Anomaly techniques intensively studied this purpose by researchers over last couple decades, which employed statistical, artificial intelligence machine learning models. Supervised algorithms most popular types models research up until recently. However, supervised associated with challenges that can be addressed semi-supervised unsupervised proposed recently published literature. This survey aims investigate present a thorough review effective anomaly applied focus on highlighting recent advancements areas learning.

Language: Английский

---

Time series forecasting of Covid-19 using deep learning models: India-USA comparative case study

Chaos Solitons & Fractals, Journal Year: 2020, Volume and Issue: 140, P. 110227 - 110227

Published: Aug. 20, 2020

Language: Английский

---

A review on deep learning in machining and tool monitoring: methods, opportunities, and challenges

The International Journal of Advanced Manufacturing Technology, Journal Year: 2021, Volume and Issue: 115(9-10), P. 2683 - 2709

Published: May 31, 2021

Language: Английский

---

Autonomous robotic laparoscopic surgery for intestinal anastomosis

Science Robotics, Journal Year: 2022, Volume and Issue: 7(62)

Published: Jan. 26, 2022

Autonomous robotic surgery has the potential to provide efficacy, safety, and consistency independent of individual surgeon's skill experience. anastomosis is a challenging soft-tissue task because it requires intricate imaging, tissue tracking, surgical planning techniques, as well precise execution via highly adaptable control strategies often in unstructured deformable environments. In laparoscopic setting, such surgeries are even more need for high maneuverability repeatability under motion vision constraints. Here we describe an enhanced autonomous strategy soft demonstrate small bowel phantom vivo intestinal tissues. This allows operator select among autonomously generated plans robot executes wide range tasks independently. We then use our perform on porcine models over 1-week survival period. compared quality criteria-including needle placement corrections, suture spacing, bite size, completion time, lumen patency, leak pressure-of developed system, manual surgery, robot-assisted (RAS). Data from model indicate that system outperforms expert surgeons' technique RAS terms accuracy. was also replicated model. These results robots exhibiting levels autonomy have improve consistency, patient outcomes, access standard technique.

Language: Английский

---