Conference Agenda

The present systematic research aims to analyze the influence of Machine Learning applications in supply chain management, focusing on studies published between 2020 and 2024. Thirty relevant researches were identified that met the established inclusion criteria. The results show that Machine Learning significantly improves demand forecasting, inventory management and logistics optimization, allowing to reduce operating costs and increase organizational efficiency. In addition, outstanding applications such as Long and Short Term Memory (LSTM) neural networks for accurate predictions and Support Vector Machines (SVM) for complex classifications were identified. However, challenges remain, such as data quality, integration with traditional systems, and the need for professionals trained in this technology. Despite these limitations, the pharmaceutical, manufacturing and food sectors have demonstrated a positive impact by implementing Machine Learning-based solutions, optimizing processes and increasing their competitiveness. In conclusion, Machine Learning is consolidating as a key driver to transform supply chain management, although more investment in training and integration strategies is required to maximize its potential.

Fraudulent transactions represent a significant global issue due to their economic and social impact. This research aims to identify the most effective machine learning models for detecting fraudulent transactions. A systematic literature review was conducted as the primary methodology, structured around three specific questions derived from the main research question: Which machine learning models are the most effective for detecting fraudulent transactions? A total of 78 articles were analyzed, extracted from the Scopus and Web of Science databases up to September 2024. Of these, 39 met the inclusion criteria established under the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) protocol. The results highlight that machine learning and deep learning models, such as Random Forest (RF), Extreme Gradient Boosting (XGB), Long Short-Term Memory (LSTM), Artificial Neural Networks (ANN), and some hybrid ML-DL models, are the most effective for detecting fraudulent transactions. It is concluded that these techniques provide robust and reliable solutions to prevent losses caused by fraud, contributing to the development of advanced strategies in organizations and financial sectors.

Artificial intelligence models have significantly transformed the Internet of Things (IoT) sector, but they have also increased the risks associated with cybersecurity. This research aims to determine which of these models are most effective in countering cybersecurity threats in IoT devices integrated into a Blockchain ecosystem. A systematic literature review was developed, structured around three specific questions derived from the main research question: What are the mosts reliable artificial intelligence models when integrated into a Blockchain ecosystem to counter cybersecurity threats in IoT devices? The analysis was performed using the PRISMA protocol and information sourced from the Scopus and Web of Science databases. The findings reveal that Machine Learning and Deep Learning models stand out in terms of effectiveness, with CNN, LSTM, and Federated Learning identified as the most reliable approaches in IoT environments within Blockchain ecosystems. Accordingly, it is concluded that these techniques provide robust and reliable solutions for mitigating risks in IoT, significantly contributing to the implementation of advanced cybersecurity strategies in organizations and technological sectors.

This project article a solution to problems with insecurity by detecting the misuse of weapons in public places, it seeks to meet the objectives of creating its own dataset to carry out training and thus be able to evaluate accuracy, precision, sensitivity and F1 Score, validating using different data sets and with these metrics the performance of the model. For the development of the project, the literature related to convolutional neural networks and weapon identification was first investigated, and cameras that allow GPS location to be extracted were also investigated. Then, a dataset made up of network images and own images taken was created. In a real scenario, the model was trained in Google Colab to identify 6 different types of weapons, including firearms and edged weapons. For this, a Yolov8 model was used, which is ideal for detection, classification and segmentation tasks. A program was created in Visual Studio Code to carry out the tests. After the tests were completed, the confusion matrices were created and the different metrics were calculated to evaluate the performance of the model in the different classes. The conclusion was reached after the analysis of the confusion matrices that the model has a medium performance when detecting images of fire (rifles, shotguns, pistols), however it has a low performance when identifying edged weapons (knives, scissors, machetes) especially in knives that are confused with other classes, especially with machetes.

This research presents the accuracy with which the NARX neural network predicts the generation of electrical energy from photovoltaic plants. The study employed a correlational design, which facilitated the description of the relationship between two variables: X = NARX Neural Networks and Y = Accuracy of Electrical Energy Generation Prediction from Photovoltaic Plants. The prediction consisted of future values of a time series of electrical energy generated by the photovoltaic plants y(t) from the past values of two time series: previously generated electrical energy and past values of solar radiation received during the same period x(t). Two cases were analyzed following this sequence: construction of the neural network, training, validation, and testing of the neural network to achieve the prediction. Finally, the prediction accuracy was evaluated through linear regression analysis, using the correlation coefficient “r” between the outputs and the targets as an indicator. This indicated how well the variation in the output was related to the targets, determining the level of accuracy.

This research performs a comparative analysis of the precision level of time series neural networks using the NARX, NAR, and input-output models for predicting the thermal energy generated by flat-plate vacuum solar collectors y(t) based on specific time series neural network models x(t). For the prediction analysis of each model, the neural network was constructed, followed by the phases of training, validation, and testing to obtain the respective predictions. The prediction level of each implemented model was then determined through linear regression analysis, which indicated how well the generated output was related to the targets. Finally, the prediction levels of the three models were compared to determine which model had a better precision for predicting the thermal energy generation of flat-plate vacuum solar collectors.