| viXra.org > Artificial Intelligence > 2005 |
 |
 |
| viXra.org > Artificial Intelligence > 2005 |
|
|
[8] viXra:2005.0160 [pdf] submitted on 2020-05-14 16:19:23
Authors: Gokhan Cagrici
Comments: 8 Pages.
Tremendous achievement of reaching fairly
high success metric values with several NLI
datasets caused eyebrows to raise questioning
the real value of these metric numbers. Research
papers started to appear with a comprehensive
analysis of what these models really
learn and the relative difficulty of forcing
these models to fail with small syntactic and
semantic changes in the input. In particular,
ANLI benchmark is an example of a more challenging
NLI task with the intent of measuring
the comprehension capabilities of models to a
deeper context.
Relative success of transformer-based models
on ANLI benchmarks were already reported
by Nie et al., 2019. Given the challenging
nature of iterative dataset formation, individual
models are having more difficulty of extracting
the underlying relationship between
the context and hypothesis pair, and the target.
Ensembles of these individual models might
have a higher potential to achieve better performance
numbers when the individual performances
are that far from the equivalent ones in
SNLI and MNLI tasks. On top of that, making
controlled variations of the inputs and tracking
the changes in the behavior of those models
will give indications about the strength and robustness
regarding the learning process.
Category: Artificial Intelligence
[7] viXra:2005.0159 [pdf] replaced on 2024-08-03 18:56:15
Authors: Teo Banica
Comments: 400 Pages.
This is an introduction to noncommutative geometry, from an affine viewpoint, that is, by using coordinates. The spaces $mathbb R^N,mathbb C^N$ have no free analogues in the operator algebra sense, but the corresponding unit spheres $S^{N-1}_mathbb R,S^{N-1}_mathbb C$ do have free analogues $S^{N-1}_{mathbb R,+},S^{N-1}_{mathbb C,+}$. There are many examples of real algebraic submanifolds $Xsubset S^{N-1}_{mathbb R,+},S^{N-1}_{mathbb C,+}$, some of which are of Riemannian flavor, coming with a Haar integration functional $int:C(X)tomathbb C$, that we will study here. We will mostly focus on free geometry, but we will discuss as well some related geometries, called easy, completing the picture formed by the 4 main geometries, namely real/complex, classical/free.
Category: Artificial Intelligence
[6] viXra:2005.0120 [pdf] submitted on 2020-05-10 13:03:10
Authors: Alexey Kutalev
Comments: 9 Pages.
Not so long ago, a method was discovered that successfully overcomes the catastrophic forgetting of neural networks. Although we know about the cases of using this method to preserve skills when adapting pre-trained networks to particular tasks, it has not yet obtained widespread distribution. In this paper, we would like to propose an alternative method of overcoming catastrophic forgetting based on the total absolute signal passing through each connection in the network. This method has a simple implementation and seems to us essentially close to the processes occurring in the brain of animals to preserve previously learned skills during subsequent learning. We hope that the ease of implementation of this method will serve its wide application.
Category: Artificial Intelligence
[5] viXra:2005.0100 [pdf] submitted on 2020-05-08 04:24:41
Authors: Parsa Rajabzadeh, Hamed Rahim
Comments: 11 Pages.
This paper aims to propose a novel MIMO control system that is compounded with Distributed Control Systems (DCS) and Centralized Control Systems (CCS). Despite DCS and CCS, which have several drawbacks such as cost and delay, the proposed system is designed to have local and global controllers simultaneously. This MIMO control system has a significant advantage versus the two traditional systems in implementation, computation power reduction, cost decrementing, performance, and the problems that occur in addressing the system connections in DCs for Wireless Sensor Networks and the Internet of Things. The proposed
the system is modeled as a Multi-Agent System (MAS) which is implemented in the osBrain MAS framework in Python.
Category: Artificial Intelligence
[4] viXra:2005.0099 [pdf] submitted on 2020-05-08 07:28:57
Authors: Kalu Kelechi Gabriel
Comments: 7 Pages.
Control systems have been in existence for quite a long time now. The oldest and unarguably the best of these is the human brain. Some popular control methodologies are PID control, Bayesian control, neural networks, etc. The drawbacks of these are however lead by a striking point that all of them are either based on Boolean or multi-valued logic, which is no more than a threshold or point-to-point logic. This work tends to introduce fuzzy logic control as a paradigm shift, this control method would to a large extent provide a physical model of the human brain. The distinction is that it is not only multi-valued logic but also a ‘degree’ based logic. This paper would give a basic overview of a fuzzy control system and its physical implementation considerations.
Category: Artificial Intelligence
[3] viXra:2005.0023 [pdf] submitted on 2020-05-02 08:30:28
Authors: Vamsi K, Ganeshan M
Comments: Pages.
The Internet is used in many situations like to send money to a friend or receive money or you might want to buy something, and we're used to using our Visa cards or PayPal or different payment methods in order to transfer money and handle using these traditional methods to transfer money is not anonymous and not private. Therefore, we'll need different methods if we want to protect our privacy and anonymity. You're probably already thinking now of using Cryptocurrencies and that's correct. Some of the cryptocurrencies are actually very secure and very anonymous. So, in this paper I am going to talk about Cryptocurrencies what it is and how it works. We're going to talk about bitcoins obviously because it's the most common Cryptocurrencies. And then we'll also go to talk about a more private Cryptocurrencies which is Monero. So, we're going to talk about how to properly obtain these Cryptocurrencies anonymously and privately how to handle them in a secure and private manner and how to send and receive. So how to transfer these currencies again in a secure private and anonymous manner. In few scenarios in case you needed to transfer money to a friend or to another person in an anonymous manner or if you wanted to pay for something if you wanted to buy something anonymously and privately or if you're simply just buying from a website that only accepts Cryptocurrencies.
Category: Artificial Intelligence
[2] viXra:2005.0003 [pdf] submitted on 2020-05-01 07:59:43
Authors: George Rajna
Comments: 38 Pages.
A paper published in Advanced Photonics "Enhanced light–matter interactions in dielectric nanostructures via machine-learning approach," suggests that machine-learning techniques can be used to enhance metasurfaces, optimizing them for nonlinear optics and optomechanics. [25]
Researchers have mathematically proven that a powerful classical machine learning algorithm should work on quantum computers. [24]
Researchers at Oregon State University have used deep learning to decipher which ribonucleic acids have the potential to encode proteins. [23]
Category: Artificial Intelligence
[1] viXra:2005.0002 [pdf] submitted on 2020-05-01 08:20:36
Authors: George Rajna
Comments: 68 Pages.
Now, SLAC researchers have developed a new tool, using machine learning, that may make part of the tuning process five times faster compared to previous methods. [40] Compared with the previous method of data pre-processing, the new machine-learning-based method has quadrupled quality metrics for the identification of particles on the calorimeter. [39] From the data collected by the LHCb detector at the Large Hadron Collider, it appears that the particles known as charm mesons and their antimatter counterparts are not produced in perfectly equal proportions. The inclusion of short-range interactions in models of neutrinoless double-beta decay could impact the interpretation of experimental searches for the elusive decay. [34] The occasional decay of neutrons into dark matter particles could solve a long-standing discrepancy in neutron decay experiments. [33] The U.S. Department of Energy has approved funding and start of construction for the SuperCDMS SNOLAB experiment, which will begin operations in the early 2020s to hunt for hypothetical dark matter particles called weakly interacting massive particles, or WIMPs. [32] Thanks to low-noise superconducting quantum amplifiers invented at the University of California, Berkeley, physicists are now embarking on the most sensitive search yet for axions, one of today's top candidates for dark matter. [31]
Category: Artificial Intelligence
| Third-Party Links: |
|