Comparative Analysis of Machine Learning Algorithms for Enhancing Social Media Marketing and Decision-Making in Kenyan SMEs.

African Journal of Commercial Studies, Journal Year: 2025, Volume and Issue: 6(1), P. 39 - 52

Published: Jan. 7, 2025

Small and medium-sized enterprises (SMEs) in Kenya are crucial to the nation's economic advancement, yet they sometimes have difficulties competing a rapidly digitalizing market due limited resources inadequate marketing strategies. Social media platforms such as Facebook, Instagram, X (formerly Twitter) essential tools for cost-effective marketing; nevertheless, many SMEs fail leverage their potential lack of data-driven strategy. Machine Learning (ML) algorithms offer transformative method examine social data, enhance campaigns, refine decision-making. This research conducts comparative analysis five prominent machine learning algorithms: Logistic Regression, Decision Trees, Random Forests, Support Vector Machines (SVM), Neural Networks, with objective improving campaigns decision-making Kenya. The researchers assess effectiveness these critical functions, including consumer segmentation, sentiment analysis, campaign optimization. A dataset comprising engagement indicators, customer profiles, performance metrics from Kenyan was used evaluate algorithms' accuracy, precision, recall, F1 score, computational efficiency. findings demonstrate that Forests strike balance between accuracy efficiency, making them feasible choice small constrained resources. Regression is suitable basic jobs, while Networks proficient at handling unstructured data but require significant computer trees, despite being understandable user-friendly, prone overfitting, whereas support vector machines, although effective datasets, large-scale applications. indicates challenges, insufficient technical expertise, elevated computing expenses, privacy issues, hinder use by It also highlights cloud-based platforms, government private sectors SME training, partnerships improve accessibility solutions. contributes growing body knowledge on application ML provides actionable recommendations harness technologies improved informed

Language: Английский

---

Use of CPT and other parameters for estimating soil unit weight using optimised machine learning models

Earth Science Informatics, Journal Year: 2025, Volume and Issue: 18(2)

Published: Jan. 31, 2025

Language: Английский

---

Large disparities in spatiotemporal distributions of building carbon emissions across China

Building and Environment, Journal Year: 2025, Volume and Issue: unknown, P. 112778 - 112778

Published: Feb. 1, 2025

Language: Английский

---

Untargeted Lipidomic Biomarkers for Liver Cancer Diagnosis: A Tree-Based Machine Learning Model Enhanced by Explainable Artificial Intelligence

Medicina, Journal Year: 2025, Volume and Issue: 61(3), P. 405 - 405

Published: Feb. 26, 2025

Background and Objectives: Liver cancer ranks among the leading causes of cancer-related mortality, necessitating development novel diagnostic methods. Deregulated lipid metabolism, a hallmark hepatocarcinogenesis, offers compelling prospects for biomarker identification. This study aims to employ explainable artificial intelligence (XAI) identify lipidomic biomarkers liver develop robust predictive model early diagnosis. Materials Methods: included 219 patients diagnosed with healthy controls. Serum samples underwent untargeted analysis LC-QTOF-MS. Lipidomic data univariate multivariate analyses, including fold change (FC), t-tests, PLS-DA, Elastic Network feature selection, significant candidate lipids. Machine learning models (AdaBoost, Random Forest, Gradient Boosting) were developed evaluated utilizing these differentiate cancer. The AUC metric was employed optimal model, whereas SHAP utilized achieve interpretability model’s decisions. Results: Notable alterations in profiles observed: decreased sphingomyelins (SM d39:2, SM d41:2) increased fatty acids (FA 14:1, FA 22:2) phosphatidylcholines (PC 34:1, PC 32:1). AdaBoost exhibited superior classification performance, achieving an 0.875. identified 40:4 as most efficacious predictions. d41:2 d36:3 lipids specifically associated risk low-onset elevated levels lipid. Conclusions: demonstrates that lipidomics, conjunction machine learning, may effectively detection results suggest metabolism are crucial progression provide valuable insights incorporating lipidomics into precision oncology.

Language: Английский

---

A predictive modelling approach to decoding consumer intention for adopting energy-efficient technologies in food supply chains

Decision Analytics Journal, Journal Year: 2025, Volume and Issue: unknown, P. 100561 - 100561

Published: March 1, 2025

Language: Английский

---

Improving smuon searches with neural networks

The European Physical Journal C, Journal Year: 2025, Volume and Issue: 85(1)

Published: Jan. 22, 2025

Abstract We demonstrate that neural networks can be used to improve search strategies, over existing in LHC searches for light electroweak-charged scalars decay a muon and heavy invisible fermion. propose new involving network discriminator as final cut show different signal regions defined using trained on subsets of samples (distinguishing low-mass high-mass regions). also present workflow publicly-available analysis tools, lead, from background simulation, training, through finding projections limits an libraries interface recasting tools. provide estimate the sensitivity our Run 2 data, higher luminosities, showing clear advantage previous methods.

Language: Английский

---

Probing sub-TeV Higgsinos aided by a machine-learning-based top tagger in the context of trilinear R -parity violating SUSY

Physical review. D/Physical review. D., Journal Year: 2025, Volume and Issue: 111(9)

Published: May 8, 2025

Probing Higgsinos remains a challenge at the LHC owing to their small production cross sections and complexity of decay modes nearly mass degenerate Higgsino states. The existing limits on are much weaker compared its bino wino counterparts. This leaves large chunk sub-TeV supersymmetric parameter space unexplored so far. In this work, we explore possibility probing masses in 400–1000 GeV range. We consider simplified scenario where R-parity is violated through baryon number violating trilinear coupling. adopt machine-learning-based top tagger tag boosted jets originating from Higgsinos, for our collider analysis, use decision tree classifier discriminate signal over SM backgrounds. construct two regions characterized by least one jet different multiplicities b-jets light jets. Combining statistical significance obtained regions, show that as high 925 can be probed HL-LHC. Published American Physical Society 2025

Language: Английский

---

Modern machine learning and particle physics: an in-depth review

The European Physical Journal Special Topics, Journal Year: 2024, Volume and Issue: 233(15-16), P. 2421 - 2424

Published: Oct. 16, 2024

Modern machine learning (ML) techniques are ubiquitous in the field of particle physics. These ML models primarily meant for exploiting large amounts high-dimensional data to reduce complexity and extract as much information possible from data. This special issue presents a series ten contributions area application modern theoretical experimental

Language: Английский

---