| viXra.org > Artificial Intelligence > 2404 |
 |
 |
| viXra.org > Artificial Intelligence > 2404 |
|
|
[6] viXra:2404.0133 [pdf] submitted on 2024-04-29 18:49:33
Authors: Budee U. zaman
Comments: 15 Pages.
The generation at the helm faces an unprecedented responsibility in the near future of artificial intelligence. The implications of setting up the founding rules that will regulate the operation of AI are heavy since after they’re set they last forever. Once this first AI is commenced, it can be such that no other subsequent AIs could emerge thereby assuming dominion over its own creation stand. As a result, retaining control becomes necessary. Lest humanity surrender agency to its own creation. At this juncture of big talks, critical issue are raised concerning AIadministration owners. Is it appropriate for only a few people to have unrestricted control on AI commands while leaving out all precautionary measure? Therefore, we have to always consider between control andconstraint when dealing with AI issues which involves authority plays off against morality. The direction Artificial Intelligence takes in the future depends on the decisions made by today’s generation. We will determinehow we are viewed historically in terms of technology based on how well we take on such an important duty. There’s a major turning point ahead of us where we who are the stewards of tomorrow must make a choice that protects humanity’s right to self-determination and also exploits the power of AI for change.
Category: Artificial Intelligence
[5] viXra:2404.0123 [pdf] submitted on 2024-04-25 15:58:34
Authors: Brady Steele
Comments: 40 Pages. CC BY: Creative Commons Attribution
This research paper presents an in-depth exploration of a neural network architecture tailored for intent classification using sentence embeddings. The model comprises a feedforward neural network with two hidden layers, ReLU activation functions, and softmax activation in the output layer. This paper meticulously examines the technical intricacies involved in data preprocessing, model architecture definition, training methodologies, and evaluation criteria. Detailed explanations are provided for the rationale behind architectural decisions, including the incorporation of dropout layers for regularization and class weight balancing techniques for handling imbalanced datasets. Moreover, the mathematical foundations of the chosen loss function (sparse categorical crossentropy) and optimization algorithm (Adam optimizer) are thoroughly elucidated, shedding light on their roles in facilitating model training and convergence. Through empirical experiments and theoretical analyses, this paper offers insights into the effectiveness and resilience of the proposed neural network architecture for intent classification tasks. It serves as a technical guide for engineers aiming to comprehend, implement, and optimize neural network models for practical application in natural language processing endeavors.
Category: Artificial Intelligence
[4] viXra:2404.0091 [pdf] submitted on 2024-04-17 20:48:34
Authors: Koffka Khan
Comments: 9 Pages. (Note by viXra Admin: Please submit article in pdf only)
As the demand for high-quality video content continues to surge, the effectiveness of adaptive video streaming hinges on the efficiency of dynamic content delivery policies. Traditional approaches face challenges in providing real-time adjustments to account for network conditions and user preferences. This review paper explores the transformative potential of blockchain technology in revolutionizing content delivery policies for adaptive streaming. We delve into the decentralized and transparent nature of blockchain to facilitate dynamic adjustments in real-time, considering factors such as network conditions and user preferences. Through an examination of existing solutions, case studies, and implementations, we showcase how blockchain can enhance the adaptive streaming experience. The paper also discusses the benefits, limitations, and future directions, providing a comprehensive overview of the role of blockchain in shaping the future of adaptive video streaming.
Category: Artificial Intelligence
[3] viXra:2404.0081 [pdf] submitted on 2024-04-15 23:43:11
Authors: Koffka Khan
Comments: 6 Pages.
In the era of big data, the exponential growth in data volume, velocity, variety, and veracity has presented unprecedented challenges for traditional data processing and analytics techniques. In response to these challenges, metaheuristic algorithms have emerged as powerful tools for solving optimization problems in large-scale datasets. This paper provides a comprehensive review of the applications of metaheuristics in addressing various challenges posed by big data. We begin with an overview of big data challenges and the characteristics of metaheuristic algorithms. We then survey the literature on the application of metaheuristics in key areas such as data preprocessing, clustering, classification, association rule mining, and optimization. Furthermore, we discuss the scalability, efficiency, adaptability, and ethical considerations associated with the use of metaheuristic algorithms in big data analytics. Finally, we outline potential directions for future research in this rapidly evolving field. This review serves as a valuable resource for researchers, practitioners, and decision-makers interested in leveraging metaheuristic approaches to extract actionable insights from big data.
Category: Artificial Intelligence
[2] viXra:2404.0075 [pdf] replaced on 2025-02-18 06:53:11
Authors: Dimiter Dobre
Comments: 16 Pages.
The goal of AI is to predict the future and use this prediction as a basis for choosing its further course of action. AI tries to understand how the world works which means that it should find a model of the world. That model consists of internal states and the function that drives transitions from one internal state to another. AI will need that model in order to predict the next observation, i.e. in order to predict the future.
For AI to gain self-awareness, it must find the answer to the questions "Where am I?" and "What is going on?". The answer to these questions is hidden in the internal state of the world. An AI which does not endeavor to understand the world is weak AI. The way to creating a strong AI goes through the description of the internal state of the world.
If we are to create Artificial General Intelligence (AGI), it would not be sufficient just to learn how to describe the internal state of the world. We also need to move from single-step to multi-step reasoning. This means that we should be able to start from the current state of the world and mentally take several steps into the future, and thereby select the course of action that works best for us.
Category: Artificial Intelligence
[1] viXra:2404.0069 [pdf] submitted on 2024-04-14 22:12:50
Authors: Ait-taleb Nabil
Comments: 11 Pages.
In the context of multiple causation, I will introduce the causation function. This function is a quadratic form computed from the correlations and serves as a generalization of R-squared, commonly found in machine learning. In this report, the causation function will make the link between the correlations and causal relationship. By examining the causation function through an illustrative example, we will demonstrate how strong or weak correlations between multiple causes and a variable can imply either a highly likely or unlikely causal relationship between the causes and the variable.
Category: Artificial Intelligence
| Third-Party Links: |
|