| viXra.org > Artificial Intelligence > 2309 |
 |
 |
| viXra.org > Artificial Intelligence > 2309 |
|
|
[5] viXra:2309.0149 [pdf] submitted on 2023-09-29 08:55:44
Authors: Farid Soroush
Comments: 12 Pages.
Machine learning has undergone tremendous advancements, paving the way for a myriad of applications across industries. In the midst of this progress, the significance of hyperparameter tuning and model evaluation can't be understated, as they play a critical role in achieving optimal model performance. This project delves into the realm of ML model optimization and evaluation, harnessing Bayesian Optimization, SHAP (SHapley Additive exPlanations), and traditional evaluation matrices. By focusing on a decision tree classifier, the study investigates the efficiency of various hyperparameter tuning methods, the interpretability of model decisions, and the robustness of performance metrics. Preliminary results suggest that Bayesian Optimization may offer advantages in efficiency over traditional tuning methods. Furthermore, SHAP values provide deeper insights into model decision-making, fostering better transparency and trust in ML applications.
Category: Artificial Intelligence
[4] viXra:2309.0107 [pdf] submitted on 2023-09-22 00:36:36
Authors: Han Ok Chol, Hyon Hui Song, Pak Chol Ryong
Comments: 9 Pages.
This paper proposes how to detect malicious network data effectivelyby the combination of sparse-response deep belief network and support vector machine.The Sparse-response Deep belief networks (SR-DBN) is an efficient non-supervised leaning machine for learning feature representation of the data without redundancy and the Support Vector Machine is designed to develop a classifier, which has high generalization ability in the feature space, in a supervised manner. In this paper, the feature representation of anomalous payload is performed by Sparse-response Deep belief Networks(SR-DBN), while the classification of normal or abnormal payload is performed by Support Vector Machine. Simulations and experiments show that the proposed abnormal network-detecting system is higher detection rate than the multi-layer perceptron which has stacked auto-encoder.
Category: Artificial Intelligence
[3] viXra:2309.0087 [pdf] submitted on 2023-09-17 15:56:13
Authors: Petar Radanliev, David De Roure, Omar Santos
Comments: 30 Pages.
In the contemporary digital age, Quantum Computing and Artificial Intelligence (AI) convergence is reshaping the cyber landscape, introducing both unprecedented opportunities and potential vulnerabilities.This research, conducted over five years, delves into the cybersecurity implications of this convergence, with a particular focus on AI/Natural Language Processing (NLP) models and quantum cryptographic protocols, notably the BB84 method and specific NIST-approved algorithms. Utilising Python and C++ as primary computational tools, the study employs a "red teaming" approach, simulating potential cyber-attacks to assess the robustness of quantum security measures. Preliminary research over 12 months laid the groundwork, which this study seeks to expand upon, aiming to translate theoretical insights into actionable, real-world cybersecurity solutions. Located at the University of Oxford's technology precinct, the research benefits from state-of-the-art infrastructure and a rich collaborative environment. The study's overarching goal is to ensure that as the digital world transitions to quantum-enhanced operations, it remains resilient against AI-driven cyber threats. The research aims to foster a safer, quantum-ready digital future through iterative testing, feedback integration, and continuous improvement. The findings are intended for broad dissemination, ensuring that the knowledge benefits academia and the global community, emphasising the responsible and secure harnessing of quantum technology.
Category: Artificial Intelligence
[2] viXra:2309.0076 [pdf] submitted on 2023-09-16 19:33:23
Authors: Nana Abeka Otoo, Muhammad Abubakar
Comments: 6 Pages.
This paper introduces Nafes as a prototype-based feature selection package designed as a wrapper centered on the highly interpretable and powerful Generalized Matrix Learning Vector Quantization (GMLVQ) classification algorithm and its local variant (LGMLVQ). Nafes utilizes the learned relevances evaluated by the mutation validation scheme for Learning Vector quantization (LVQ), which iteratively converges to selected features that relevantly contribute to the prototype-based classifier decisions.
Category: Artificial Intelligence
[1] viXra:2309.0063 [pdf] submitted on 2023-09-12 04:24:58
Authors: Hernández Rodríguez, Matías Ezequiel
Comments: 10 pages, 2 figures
In this article, we propose a new metaheuristic inspired by the morphogenetic cellular movements of endothelial cells (ECs) that occur during the tumor angiogenesis process. This algorithm starts with a random initial population. In each iteration, the best candidate selected as the tumor, while the other individuals in the population are treated as ECs migrating toward the tumor's direction following a coordinated dynamics through a spatial relationship between tip and follower ECs. EC movements mathematical model in angiogenic morphogenesis are detailed in the article.This algorithm has an advantage compared to other similar optimization metaheuristics:the model parameters are already configured according to the tumor angiogenesis phenomenon modeling, preventing researchers from initializing them with arbitrary values.Subsequently, the algorithm is compared against well-known benchmark functions, and the results are validated through a comparative study with Particle Swarm Optimization (PSO). The results demonstrate that the algorithm is capable of providing highly competitive outcomes.Also the proposed algorithm is applied to a real-world problem. The results showed that the proposed algorithm performed effective in solving constrained optimization problems, surpassing other known algorithms.
Category: Artificial Intelligence
| Third-Party Links: |
|