scholarly journals Evolutionary algorithms with and without adaptive mutation in AI based cryptography

2018 ◽  
Vol 21 ◽  
pp. 00008
Author(s):  
Mateusz Tybura
Keyword(s):  
Neural Networks ◽  
Evolutionary Algorithms ◽  
Evolutionary Algorithm ◽  
Transfer Learning ◽  
Adaptive Mutation ◽  
Cryptographic Keys ◽  
Similar Task

The key role of cryptography is to make cipher so hard to reproduce without knowing all the details that no one besides the recipient could decipher the message. Those algorithms which are used nowadays gets its security mostly from highly reliable algorithms and/or complicated cryptographic keys. Unfortunately, those human-made methods aren‘t invulnerable so sooner or later they compromise. So, it could be really useful to make a cipher which could change. But currently only neural networks are capable of thing known as transfer learning. In this article similar method was proposed in order to make it possible to re-learn already established evolutionary algorithm to do new, similar task.

scholarly journals On the Choice of the Offspring Population Size in Evolutionary Algorithms

2005 ◽  
Vol 13 (4) ◽  
pp. 413-440 ◽  
Author(s):  
Thomas Jansen ◽  
Kenneth A. De Jong ◽  
Ingo Wegener
Keyword(s):  
Evolutionary Algorithms ◽  
Evolutionary Algorithm ◽  
Population Size ◽  
Diffi Cult ◽  
Very High

Evolutionary algorithms (EAs) generally come with a large number of parameters that have to be set before the algorithm can be used. Finding appropriate settings is a diffi- cult task. The influence of these parameters on the efficiency of the search performed by an evolutionary algorithm can be very high. But there is still a lack of theoretically justified guidelines to help the practitioner find good values for these parameters. One such parameter is the offspring population size. Using a simplified but still realistic evolutionary algorithm, a thorough analysis of the effects of the offspring population size is presented. The result is a much better understanding of the role of offspring population size in an EA and suggests a simple way to dynamically adapt this parameter when necessary.

scholarly journals QUANTUM-INSPIRED ARTIFICIL NEURAL NETWORKS AND EVOLUTIONARY ALGORITHMS METHODS APPLIED TO MODELING OF THE POLISH ELECTRIC POWER EXCHANGE USING THE DAY-AHEAD MARKET DATA

2018 ◽  
Vol 7 (3) ◽  
pp. 201-212
Author(s):  
Jerzy Tchórzewski ◽  
Dariusz Ruciński
Keyword(s):  
Artificial Intelligence ◽  
Neural Networks ◽  
Quantum Computing ◽  
Evolutionary Algorithms ◽  
Evolutionary Algorithm ◽  
Computer Science ◽  
Electric Power ◽  
Quantum Computer ◽  
Neural Model ◽  
Power Exchange

The paper presents selected results of research on the use of artificial intelligence methods, which are inspired by quantum computing solutions for modelling of electric power exchange systems. Methods used in the modelling of quantum data acquisition, quantization and dequantization of information as well as the methods of performing quantum computations were emphasized. Furthermore, we have analysed the results obtained for the neural model and for the evolutionary algorithm inspired by the quantum computer science. Eventually, the model was verified on the example of the neural model of the Electric Power Exchange (EPE).

Technique of scientific and technical forecasting using neural networks

2018 ◽  
pp. 103-110
Author(s):  
V. V. Nefedev
Keyword(s):  
Mathematical Modeling ◽  
Neural Networks ◽  
Artificial Neural Networks ◽  
Original Method ◽  
Forecasting Methods ◽  
Expert Assessments ◽  
Artificial Neural

For the definition and implementation of breakthrough technologies the most important is the role of scientific and technical forecasting. Well-known forecasting methods based on extrapolation, expert assessments and mathematical modeling are not universal and have a number of significant disadvantages. The article proposes an original method of scientific and technical forecasting based on the use of the methodology of artificial neural networks. 

scholarly journals Interpretation of Swedish Sign Language Using Convolutional Neural Networks and Transfer Learning

2021 ◽  
Vol 2 (3) ◽  
Author(s):  
Gustaf Halvardsson ◽  
Johanna Peterson ◽  
César Soto-Valero ◽  
Benoit Baudry
Keyword(s):  
Neural Networks ◽  
Sign Language ◽  
Transfer Learning ◽  
Convolutional Neural Networks ◽  
Web Application ◽  
Training Dataset ◽  
Motion Processing ◽  
Image Perception ◽  
Sign Languages ◽  
High Level

AbstractThe automatic interpretation of sign languages is a challenging task, as it requires the usage of high-level vision and high-level motion processing systems for providing accurate image perception. In this paper, we use Convolutional Neural Networks (CNNs) and transfer learning to make computers able to interpret signs of the Swedish Sign Language (SSL) hand alphabet. Our model consists of the implementation of a pre-trained InceptionV3 network, and the usage of the mini-batch gradient descent optimization algorithm. We rely on transfer learning during the pre-training of the model and its data. The final accuracy of the model, based on 8 study subjects and 9400 images, is 85%. Our results indicate that the usage of CNNs is a promising approach to interpret sign languages, and transfer learning can be used to achieve high testing accuracy despite using a small training dataset. Furthermore, we describe the implementation details of our model to interpret signs as a user-friendly web application.

scholarly journals Improving Semi-Supervised Learning for Audio Classification with FixMatch

Electronics ◽  
2021 ◽  
Vol 10 (15) ◽  
pp. 1807
Author(s):  
Sascha Grollmisch ◽  
Estefanía Cano
Keyword(s):  
Neural Networks ◽  
Supervised Learning ◽  
Transfer Learning ◽  
Data Transfer ◽  
State Of The Art ◽  
Training Data ◽  
Audio Classification ◽  
Image Domain ◽  
Full Dataset ◽  
Audio Data

Including unlabeled data in the training process of neural networks using Semi-Supervised Learning (SSL) has shown impressive results in the image domain, where state-of-the-art results were obtained with only a fraction of the labeled data. The commonality between recent SSL methods is that they strongly rely on the augmentation of unannotated data. This is vastly unexplored for audio data. In this work, SSL using the state-of-the-art FixMatch approach is evaluated on three audio classification tasks, including music, industrial sounds, and acoustic scenes. The performance of FixMatch is compared to Convolutional Neural Networks (CNN) trained from scratch, Transfer Learning, and SSL using the Mean Teacher approach. Additionally, a simple yet effective approach for selecting suitable augmentation methods for FixMatch is introduced. FixMatch with the proposed modifications always outperformed Mean Teacher and the CNNs trained from scratch. For the industrial sounds and music datasets, the CNN baseline performance using the full dataset was reached with less than 5% of the initial training data, demonstrating the potential of recent SSL methods for audio data. Transfer Learning outperformed FixMatch only for the most challenging dataset from acoustic scene classification, showing that there is still room for improvement.

The Role of Regularization in Overparameterized Neural Networks*

2020 ◽  
Author(s):  
Siddhartha Satpathi ◽  
Harsh Gupta ◽  
Shiyu Liang ◽  
R Srikant
Keyword(s):  
Neural Networks

scholarly journals Large-Scale Adaptive Hitchhiking Upon High Recombination inDrosophila simulans

Genetics ◽  
2003 ◽  
Vol 165 (2) ◽  
pp. 895-900 ◽  
Author(s):  
Humberto Quesada ◽  
Ursula E M Ramírez ◽  
Julio Rozas ◽  
Montserrat Aguadé
Keyword(s):  
Genetic Linkage ◽  
Large Scale ◽  
Intermediate Frequency ◽  
Core Region ◽  
Adaptive Mutation ◽  
Nucleotide Variation ◽  
Extended Haplotype ◽  
Major Haplotype ◽  
Pattern Of Variation

AbstractNatural selection is expected to leave a characteristic footprint on neighboring nucleotide variation through the effects of genetic linkage. The size of the region affected is proportional to the strength of selection and greatly reduced with the recombinational distance from the selected site. Thus, the genomic footprint of selection is generally believed to be restricted to a small DNA stretch in normal and highly recombining regions. Here, we study the effect of selection on linked polymorphism (hitchhiking effect) by surveying nucleotide variation across a highly recombining ∼88-kb genomic fragment in an African population of Drosophila simulans. We find a core region of up to 38 kb with a major haplotype at intermediate frequency. The extended haplotype structure that gradually vanishes until disappearing is unusual for a highly recombining region. Both the presence in the structured genomic domain of a single major haplotype depleted of variability and the detected spatial pattern of variation along the ∼88-kb fragment are incompatible with neutral predictions in a panmictic population. A major role of demographic effects could also be discarded. The observed pattern of variation clearly provides evidence that directional selection has acted recently on this region, sweeping out variation around a strongly adaptive mutation. Our findings suggest a major role of positive selection in shaping DNA variability even in highly recombining regions.

Sign in / Sign up

Export Citation Format

Share Document