Analysing Influential Factors in Student Academic Achievement: Prediction Modelling and Insight

International Journal of Emerging Multidisciplinaries Computer Science & Artificial Intelligence, Journal Year: 2023, Volume and Issue: 2(1)

Published: Nov. 25, 2023

The fascination with understanding student academic performance has drawn widespread attention from various stakeholders, including parents, policymakers, and businesses. 'Students Performance in Exams' dataset, available on platforms like Kaggle, stands as a treasure trove. It extends beyond test scores, encompassing diverse attributes ethnicity, gender, parental education, preparation, even lunch type. In our tech-driven age, predicting success become compelling pursuit. This study aims to delve deep into this utilizing data mining methods robust classification algorithms Logistic Regression Random Forest Jupyter Notebook environment. Rigorous model training, testing, fine-tuning strive for the utmost predictive accuracy. Data cleaning preprocessing play crucial role establishing reliable dataset accurate predictions. Beyond numbers, project emphasizes visualization's impact, transforming raw comprehensible insights effective communication. Model exhibits an impressive 87.6% accuracy, highlighting its potential performance. Moreover, excels remarkable 100% accuracy forecasting grades, showcasing effectiveness domain.

Language: Английский

---

Machine learning security and privacy: a review of threats and countermeasures

EURASIP Journal on Information Security, Journal Year: 2024, Volume and Issue: 2024(1)

Published: April 23, 2024

Abstract Machine learning has become prevalent in transforming diverse aspects of our daily lives through intelligent digital solutions. Advanced disease diagnosis, autonomous vehicular systems, and automated threat detection triage are some prominent use cases. Furthermore, the increasing machine critical national infrastructures such as smart grids, transport, natural resources makes it an attractive target for adversaries. The to systems is aggravated due ability mal-actors reverse engineer publicly available models, gaining insight into algorithms underpinning these models. Focusing on landscape we have conducted in-depth analysis critically examine security privacy threats factors involved developing adversarial attacks. Our highlighted that feature engineering, model architecture, targeted system knowledge crucial formulating one successful attack can lead other attacks; instance, poisoning attacks membership inference backdoor We also reviewed literature concerning methods techniques mitigate whilst identifying their limitations including data sanitization, training, differential privacy. Cleaning sanitizing datasets may challenges, underfitting affecting performance, whereas does not completely preserve model’s Leveraging surfaces mitigation techniques, identify potential research directions improve trustworthiness systems.

Language: Английский

---

Exploring security and privacy enhancement technologies in the Internet of Things: A comprehensive review

Security and Privacy, Journal Year: 2024, Volume and Issue: 7(6)

Published: July 12, 2024

Abstract In the era heavily influenced by Internet of Things (IoT), prioritizing strong security and protection user privacy is utmost importance. This comprehensive review paper embarks on a meticulous examination multifaceted challenges risks facing IoT privacy. It encompasses hardware, software, data‐in‐transit domains, shedding light potential vulnerabilities associated threats. response to these concerns, this puts forth recommendations for effective strategies mitigate risks. Providing road‐map enhancing in environments. Furthermore, thoroughly assesses multitude solutions proposed various authors, with primary aim within landscape. The analysis provides insights into strengths limitations solutions. aiding development holistic comprehension existing status Moreover, delves complexities surrounding integrating emerging technologies framework. explores obstacles inherent process proposes address hurdles. By doing so, perspective enhancement offers guidance navigating dynamic landscape domain. Publications included consist journal articles, conference papers, book chapters from reputable sources indexed SCI (Science Citation Index), Scopus, Web Science.

Language: Английский

---

Risk Management and Cybersecurity in Transportation and Warehousing

Advances in logistics, operations, and management science book series, Journal Year: 2023, Volume and Issue: unknown, P. 1 - 35

Published: Dec. 29, 2023

Transportation and warehousing are vital components of logistics corporations. Their continuous uninterrupted functioning is paramount significance for the enterprises involved in supply chain. As these built rely heavily on digital technologies their rapid functioning, they vulnerable to cybersecurity threats attacks. Hence, effectively address promptly respond issues organizations need have proper strategy planning place. This chapter endeavors acquaint readers with pressing issues. To secure operations transportation systems, methods tools assessing risks mitigating them discussed comprehensively.

Language: Английский

---

Empirical Performance Analysis of Hyperledger Fabric

Published: Jan. 1, 2025

Language: Английский

---

Evaluating Climate Change Impacts on Streamflow Changes in the Source Region of Yellow River: A Bayesian Vine Copula Machine Learning (BVC‐ML) Approach

International Journal of Climatology, Journal Year: 2025, Volume and Issue: unknown

Published: March 11, 2025

ABSTRACT In this study, we proposed a Bayesian Vine Copula Machine Learning (BVC‐ML) method to predict streamflow changes in the Yellow River source area based on projections from three GCMs under various climate change scenarios. The BVC‐ML was (i) use vine copula reflect interdependence between predicted variable (i.e., streamflow) and predictions different machine learning (ML) techniques, (ii) derive deterministic probabilistic model conditional corresponding ML (iii) integrate models generate final results. then applied for future outputs CMIP6. results show that studied would generally experience more increases most months, become significant as shifts SSP126 SSP585. GCM also lead area, with ACCESS‐CM2 leading highest increases. Furthermore, is capable of deriving both distributions, 10% 90% quantiles can predictive uncertainties. quantile May, July October have streamflow, which are consistent mean Overall, demonstrated be promising tool predicting findings implications water resource management adaptation over region.

Language: Английский

---

Smart home system using blockchain technology in green lighting environment in rural areas

Heliyon, Journal Year: 2024, Volume and Issue: 10(4), P. e26620 - e26620

Published: Feb. 1, 2024

Currently, with the rapid development of smart home technology, demand for establishing efficient and sustainable systems in rural areas is increasing. However, environments, effective management intelligent control green energy face many challenges. To address these issues, this work aims to design a system based on blockchain technology achieve lighting environment areas. The main goals include improving performance safety meet needs promote development. comprises two primary components: gateway cloud services. These components encompass functions like data monitoring transmission, storage, remote control. also introduces structural interaction, user node security transmission scheme system. Ultimately, system's effectiveness confirmed through simulation experiments. results demonstrate that achieves lowest latency when transaction arrival rate 40tps block size 10. Additionally, access Hyperledger Fabric consortium chain can efficiently handle requests resources practical application requirements within an appropriate range parameters. research conclusion designed has achieved significant security. This not only provides reliable solutions areas, but important theoretical guidance future systems. direction includes further optimizing performance, expanding scope application, exploring more advanced applications field homes. will provide possibilities innovative directions

Language: Английский

---

AI and Blockchain-Assisted Secure Data-Exchange Framework for Smart Home Systems

Mathematics, Journal Year: 2023, Volume and Issue: 11(19), P. 4062 - 4062

Published: Sept. 25, 2023

The rapid expansion of the Internet Things (IoT) on a global scale has facilitated convergence revolutionary technologies such as artificial intelligence (AI), blockchain, and cloud computing. integration these paved way for development intricate infrastructures, smart homes, cities, industries, that are capable delivering advanced solutions enhancing human living standards. Nevertheless, IoT devices, while providing effective connectivity convenience, often rely traditional network interfaces can be vulnerable to exploitation by adversaries. If not properly secured updated, legacy communication protocols expose potential vulnerabilities attackers may exploit gain unauthorized access, disrupt operations, or compromise sensitive data. To overcome security challenges associated with home systems, we have devised robust framework leverages capabilities both AI blockchain technology. proposed employs standard dataset from which first eliminated anomalies using an isolation forest (IF) algorithm random partitioning, path length, anomaly score calculation, thresholding stages. Next, is utilized training classification algorithms, K-nearest neighbors (KNN), support vector machine (SVM), linear discriminate analysis (LDA), quadratic discriminant (QDA) classify attack non-attack data system. Further, interplanetary file system (IPFS) store classified (non-attack data) algorithms confront data-manipulation attacks. IPFS acts onsite storage system, securely storing data, its computed hash forwarded blockchain’s immutable ledger. We evaluated different performance parameters. These include accuracy (99.53%) KNN 99.27% IF detection. used validation curve, lift execution cost transactions, scalability (86.23%) showcase effectiveness framework.

Language: Английский

---

Design of a Decentralized Identifier-Based Authentication and Access Control Model for Smart Homes

Electronics, Journal Year: 2023, Volume and Issue: 12(15), P. 3334 - 3334

Published: Aug. 4, 2023

In recent years, smart homes have garnered extensive attention as a prominent application scenario of IoT technology. However, the unique characteristics brought forth serious security threats, emphasizing paramount importance identity authentication and access control. The conventional centralized approach is plagued by issue having “single point failure,” while existing distributed solutions are constrained limited device resources complexities authentication. To tackle these challenges, this paper proposes home control model based on decentralized identifiers (DIDs). By leveraging inherent decentralization DIDs, which rely blockchain, environment constructed, effectively mitigating problem failure.” model, every participant in system, including users devices, uniquely identified DIDs through integration an improved capability-based scheme, streamlines user process, reduces complexity, enables convenient cross-household with single registration. Our experimental results demonstrate that provides various attributes, confidentiality, integrity, traceability. Additionally, exhibits low time costs for each module, ensuring timely responses to service requests incurring lower gas consumption compared other Ethereum-based methods. Thus, our research lightweight solution suitable environments.

Language: Английский

---

Heart disease Prediction Using Data Mining Techniques

Published: Jan. 8, 2024

Heart sickness is known as one of the leading causes loss life in globe. Medical tools and various hospital programs have a large amount clinical information. Therefore, understanding heart data critical to improving predictive accuracy. In wildcat analysis, 10 feature selection strategies, namely, ANOVA, Chi-square, aggregated data, Help, advanced characteristic selection, background, full choice, algorithmic removal, Lasso retreat, Ridge 6 stages, fence tree, random forest, vector support machine, K-neighbor, providing retrospect, mathematician naive Bayes, using Cleveland database cardiopathy. 88.52%, 91.30% accuracy, 80.76% sensitivity, 85-f-measure. 71% according selected tree category.

Language: Английский

---