Re-Thinking Data Strategy and Integration for Artificial Intelligence: Concepts, Opportunities, and Challenges

Applied Sciences, Год журнала: 2023, Номер 13(12), С. 7082 - 7082

Опубликована: Июнь 13, 2023

The use of artificial intelligence (AI) is becoming more prevalent across industries such as healthcare, finance, and transportation. Artificial based on the analysis large datasets requires a continuous supply high-quality data. However, using data for AI not without challenges. This paper comprehensively reviews critically examines challenges AI, including quality, volume, privacy security, bias fairness, interpretability explainability, ethical concerns, technical expertise skills. these in detail offers recommendations how companies organizations can address them. By understanding addressing challenges, harness power to make smarter decisions gain competitive advantage digital age. It expected, since this review article provides discusses various strategies over last decade, that it will be very helpful scientific research community create new novel ideas rethink our approaches AI.

Язык: Английский

---

Applications and Techniques of Machine Learning in Cancer Classification: A Systematic Review

Human-Centric Intelligent Systems, Год журнала: 2023, Номер 3(4), С. 588 - 615

Опубликована: Сен. 11, 2023

Abstract The domain of Machine learning has experienced Substantial advancement and development. Recently, showcasing a Broad spectrum uses like Computational linguistics, image identification, autonomous systems. With the increasing demand for intelligent systems, it become crucial to comprehend different categories machine acquiring knowledge systems along with their applications in present world. This paper presents actual use cases learning, including cancer classification, how algorithms have been implemented on medical data categorize diverse forms anticipate outcomes. also discusses supervised, unsupervised, reinforcement highlighting benefits disadvantages each category intelligence system. conclusions this systematic study methods classification numerous implications. main lesson is that through accurate kinds, patient outcome prediction, identification possible therapeutic targets, holds enormous potential improving diagnosis therapy. review offers readers broad understanding as advancements applied today, empowering them decide themselves whether these clinical settings. Lastly, wraps up by engaging discussion future new types be developed field advances. Overall, information included survey article useful scholars, practitioners, individuals interested gaining about fundamentals its various areas activities.

Язык: Английский

---

Modified Genetic Algorithm with Deep Learning for Fraud Transactions of Ethereum Smart Contract

Applied Sciences, Год журнала: 2023, Номер 13(2), С. 697 - 697

Опубликована: Янв. 4, 2023

Recently, the Ethereum smart contracts have seen a surge in interest from scientific community and new commercial uses. However, as online trade expands, other fraudulent practices—including phishing, bribery, money laundering—emerge significant challenges to security. This study is useful for reliably detecting transactions; this work developed deep learning model using unique metaheuristic optimization strategy. The method overcome challenges, Optimized Genetic Algorithm-Cuckoo Search (GA-CS), combined with learning. In research, Algorithm (GA) used phase of exploration Cuckoo (CS) technique address deficiency CS. A comprehensive experiment was conducted appraise efficiency performance suggested strategies compared those various popular techniques, such k-nearest neighbors (KNN), logistic regression (LR), multi-layer perceptron (MLP), XGBoost, light gradient boosting machine (LGBM), random forest (RF), support vector classification (SVC), terms restricted features we their metrics approach behavior on Ethereum. SVC models outperform rest models, highest accuracy, while proposed strategy outperforms RF model, slightly higher 99.71% versus 98.33%.

Язык: Английский

---

A Review on Nature-Inspired Algorithms for Cancer Disease Prediction and Classification

Mathematics, Год журнала: 2023, Номер 11(5), С. 1081 - 1081

Опубликована: Фев. 21, 2023

In the era of healthcare and its related research fields, dimensionality problem high-dimensional data is a massive challenge as it crucial to identify significant genes while conducting on diseases like cancer. As result, studying new Machine Learning (ML) techniques for raw gene expression biomedical an important field research. Disease detection, sample classification, early disease prediction are all analyses in bioinformatics. Recently, machine-learning have dramatically improved analysis high-dimension sets. Nonetheless, researchers’ studies faced vast dimensions, i.e., features (genes) with very low space. this paper, two-dimensionality reduction methods, feature selection, extraction introduced systematic comparison several dimension data. We presented review some most popular nature-inspired algorithms analyzed them. The paper mainly focused original principles behind each their applications cancer classification from Lastly, advantages disadvantages evaluated. This may guide researchers choose effective algorithm satisfactory

Язык: Английский

---

Re-think Data Strategy and Integration for Artificial Intelligence: Concepts, Opportunities and Challenges

Опубликована: Июнь 1, 2023

The use of artificial intelligence (AI) is becoming more prevalent across industries as diverse healthcare, finance, and transportation. Artificial based on the analysis large data sets requires a continuous supply high-quality data. However, using for AI not without its challenges. This paper comprehensively reviews critically examine challenges AI, including quality, volume, privacy security, bias fairness, interpretability explainability, ethical concern, technical expertise skills. examines e these in details offers advices how companies can address them. By understanding addressing challenges, organizations harness power to make smarter decisions gain competitive advantage digital age.

Язык: Английский

---

Optimizing Gene Selection and Cancer Classification with Hybrid Sine Cosine and Cuckoo Search Algorithm

Journal of Medical Systems, Год журнала: 2024, Номер 48(1)

Опубликована: Янв. 9, 2024

Язык: Английский

---

Novel Cuckoo Search-Based Metaheuristic Approach for Deep Learning Prediction of Depression

Applied Sciences, Год журнала: 2023, Номер 13(9), С. 5322 - 5322

Опубликована: Апрель 24, 2023

Depression is a common illness worldwide with doubtless severe implications. Due to the absence of early identification and treatment for depression, millions individuals suffer from mental illnesses. It might be difficult identify those who are experiencing health illnesses provide them help that they need. Additionally, depression may associated thoughts suicide. Currently, there no clinically specific diagnostic biomarkers can severity type depression. In this research paper, novel particle swarm-cuckoo search (PS-CS) optimization algorithm proposed instead traditional backpropagation training deep neural networks. The widely used supervised learning in networks, but it has limitations terms convergence speed possibility getting trapped local optima. These problems were addressed by using network architecture detection tasks along PS-CS technique. combines strengths both swarm cuckoo algorithms, which allows more efficient effective parameters. We also evaluated how well suggested methods performed against most classification models, including (K-nearest neighbor) KNN, (support vector regression) SVR, decision trees, as residual (ResNet), visual geometry group (VGG), simple (LeNet). findings show method, PS-CS, conjunction CNN model, outperformed all other achieving maximum accuracy 99.5%. Other such logistic regression, achieved lower accuracies ranging 69% 97%.

Язык: Английский

---

AI-driven aquaculture: A review of technological innovations and their sustainable impacts

Artificial Intelligence in Agriculture, Год журнала: 2025, Номер unknown

Опубликована: Фев. 1, 2025

Язык: Английский

---

A Review of Posture Detection Methods for Pigs Using Deep Learning

Applied Sciences, Год журнала: 2023, Номер 13(12), С. 6997 - 6997

Опубликована: Июнь 9, 2023

Analysis of pig posture is significant for improving the welfare and yield captive pigs under different conditions. Detection postures, such as standing, lateral lying, sternal sitting, can facilitate a comprehensive assessment psychological physiological conditions pigs, prediction their abnormal or detrimental behavior, evaluation farming to improve yield. With introduction smart into industry, effective applicable detection methods become indispensable realizing above purposes in an intelligent automatic manner. From early manual modeling traditional machine vision, then deep learning, multifarious have been proposed meet practical demand. Posture based on learning show great superiority terms performance (such accuracy, speed, robustness) feasibility simplicity universality) compared with most methods. It promising popularize technology actual commercial production large scale automate monitoring. This review comprehensively introduces data acquisition sub-tasks technological evolutionary processes, also summarizes application mainstream models detection. Finally, limitations current future directions research will be discussed.

Язык: Английский

---

CO‐WOA: Novel Optimization Approach for Deep Learning Classification of Fish Image

Chemistry & Biodiversity, Год журнала: 2023, Номер 20(8)

Опубликована: Июль 3, 2023

The most significant groupings of cold-blooded creatures are the fish family. It is crucial to recognize and categorize species since various seafood diseases decay exhibit different symptoms. Systems based on enhanced deep learning can replace area's currently cumbersome sluggish traditional approaches. Although it seems straightforward, classifying images a complex procedure. In addition, scientific study population distribution geographic patterns important for advancing field's present advancements. goal proposed work identify best performing strategy using cutting-edge computer vision, Chaotic Oppositional Based Whale Optimization Algorithm (CO-WOA), data mining techniques. Performance comparisons with leading models, such as Convolutional Neural Networks (CNN) VGG-19, made confirm applicability suggested method. feature extraction approach Proposed Deep Learning Model was used in research, yielding accuracy rates 100 %. performance also compared image processing models an 98.48 %, 98.58 99.04 98.44 99.18 % 99.63 Networks, ResNet150V2, DenseNet, Visual Geometry Group-19, Inception V3, Xception. Using empirical method leveraging artificial neural networks, model shown be model.

Язык: Английский

---