: Accurate demand forecasting of railway freight car components is critical for effective material planning under condition-based maintenance (CBM). Traditional forecasting methods often fail to capture nonlinear patterns and perform poorly with small and uncertain datasets. This paper proposes a demand forecasting model that integrates grey relational analysis with neural networks to improve prediction accuracy for key railway components. Based on historical consumption, maintenance, and market data from a major Chinese railway equipment company, influencing factors were first identified using grey correlation analysis. The selected features were then input into a grey neural network model to predict component demand. Comparative experiments show that the proposed model significantly outperforms traditional grey prediction and BP neural network approaches, with reductions in mean squared error and mean absolute error across multiple component types. The results demonstrate that grey neural networks can effectively handle small-sample, uncertain data and provide more reliable demand forecasts for CBM-driven railway operations. This study contributes to intelligent material management in railway enterprises and provides a practical reference for improving forecasting systems in complex industrial environments.
TL;DR: An overview of convolutional neural network-based static malware analysis techniques acknowledges the recent trend of conceptualizing malware as a sequential structure with both local and long-term dependencies, the need to reconsider the notion of dataset balance, and the need for consistent and transparent application of the F1-score.
: This paper provides an overview of convolutional neural network-based static malware analysis techniques. Three research questions are considered: Which architectures based on or related to CNNs are used in static malware analysis? Which datasets are used to support research in…
TL;DR: A conditional generative adversarial networks method that integrates the machine learning with the deep learning to detect the hardware Trojans injected in Register-Transfer Level code and it contributes to enhancing the security and trustworthiness of ICs against hardware Trojan attacks.
: Hardware Trojan (HT) can compromise the security of a system by changing the integrated circuit (IC) functionality and reducing the system ꞌ s reliability. To handle this issue, machine learning has been widely used to analyze the datasets extracted from circuits to detect hard…
TL;DR: A Reliable Resource Placement with Migration Function (MF) method to reduce the outage in SC communications is proposed and reduces outage time by 13.79%, network overload by 14.04% and improves the response ratio by 13.41% for the maximum network load.
: Smart City (SC) development with technological aspects depends on wireless communication and intelligent networks such as the Internet of Things (IoT). Wireless networks and IoT interconnect resources and projects them to be ubiquitous for various applications and user services…
TL;DR: A new intrusion detecting framework is presented in this paper that is based on a combination of a Domain-adaptive Gated Deep Belief Network (DomG-DeNet) and an enhanced optimization method known as Builder-on-Zebra Recurrent Dropout Optimization (BoZ-RDO).
: Due to the rapid growth of the modern network infrastructures and the rise in the sophistication of the attacks by criminals based on networks, intrusion detection system (IDS) has become a crucial component in offering network security. The common machine learning and the exis…
TL;DR: A comparison of dataset versions with and without the entropy feature showed that the proposed entropy calculation method improves classification performance, even though the number of features was reduced compared to the original dataset.
: In machine learning and classification, entropy holds significant potential. This paper introduces a method to calculate Shannon entropy across all features within individual records in four IDS datasets: CSE-CIC-IDS2018, CIC-IDS2017, UNSW-NB15, and LUFlow. Each dataset is resh…
TL;DR: An Adaptive Binary Genetic Algorithm (A-BGA) is developed that introduces population-diversity-driven dynamic crossover and mutation rates, and reformulates the fitness as a bi-objective trade-off between prediction RMSE and feature cardinality to form a comprehensive framework that enhances prediction efficiency and uncertainty modeling.
: As the share of renewable energy in power systems continues to grow, improving prediction accuracy has become critical for enhancing system flexibility and reducing operational costs. In this paper, we propose two novel optimization methods tailored for renewable energy predict…