Modified immune network algorithm based on the Random Forest approach for the complex objects control

Samigulina, Galina; Samigulina, Zarina

doi:10.1007/s10462-018-9621-7

Cited by 24 publications

(7 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A formal peptide is understood as a mathematical abstraction of the free energy of a protein molecule from its spatial form. This is a relatively new direction in artificial intelligence, using which a number of applications have been developed [ 31 , 43 ]. The main problems that arise during immune network modeling are: the choice of the structure of the immune network; reduction of training time; solving the problem of informative features selection; increase the reliability of the prediction and parallelization of computational algorithms.…”

Section: Methods and Algorithms Of Researchmentioning

confidence: 99%

“…The minimum value of the binding energy determines the class to which this pattern belongs to. Next, the energy errors are estimated by homologs [ 43 ]. Then there is conducted the prognosis and selection of drug candidate compounds.…”

Section: Methods and Algorithms Of Researchmentioning

confidence: 99%

“…Below there is a comparative analysis of the effectiveness of various algorithms, which showed that the AIS models after the reduction of non-informative descriptors based on GWO give the best prognostic results: GWO-AIRS 75.3%, GWO-CLONALG 76.5% and the most effective modified algorithm is GWO - AIS 93.5% [ 43 ].…”

Section: Methods and Algorithms Of Researchmentioning

confidence: 99%

See 2 more Smart Citations

Ontological model of multi-agent Smart-system for predicting drug properties based on modified algorithms of artificial immune systems

Samigulina

2020

Theor Biol Med Model

View full text Add to dashboard Cite

Background Currently, due to the huge progress in the field of information technologies and computer equipment, it is important to use modern approaches of artificial intelligence in order to process extensive chemical information at creating new drugs with desired properties. The interdisciplinary of research creates additional difficulties in creating new drugs. Currently, there are no universal algorithms and software for predicting the “structure-property” dependence of drug compounds that can take into account the needs of specialists in this field. In this regard, the development of a modern Smart-system based on the promising bio-inspired approach of artificial immune systems for predicting the structure-property dependence of drug compounds is relevant. The aim of this work is to develop a multi-agent Smart-system for predicting the “structure-property” dependence of drug compounds using the ontological approach and modified algorithms of artificial immune systems using the example of drug compounds of the sulfonamide group. The proposed system makes it possible to increase the accuracy of prediction models of the “structure-property” dependence, to reduce the time and financial costs for obtaining candidate drug compounds. Methods During the creation of a Smart-system, there are used multi-agent and ontological approaches, which allow to structure input and output data, optimally to distribute computing resources and to coordinate the work of the system. As a promising approach for processing a large amount of chemical information, extracting informative descriptors and for the creation of an optimal data set, as well as further predicting the properties of medicinal compounds, there are considered modified algorithms of artificial immune systems and various algorithms of artificial intelligence. Results There was developed an ontological model of a multi-agent Smart-system. There are presented the results of the «structure-property» dependence simulation based on a modified grey wolf optimization algorithm and artificial immune systems. During the simulation, there was used information from the Mol-Instincts sulfonamide descriptor database. Conclusion The developed multi-agent Smart-system using ontological models allows visually to present the structure and interrelationships of agents functioning, which greatly facilitates the development of software and reduces time and financial costs during the development of new drugs.

show abstract

Section: Methods and Algorithms Of Researchmentioning

confidence: 99%

Section: Methods and Algorithms Of Researchmentioning

confidence: 99%

See 1 more Smart Citation

Ontological model of multi-agent Smart-system for predicting drug properties based on modified algorithms of artificial immune systems

Samigulina

2020

Theor Biol Med Model

View full text Add to dashboard Cite

show abstract

“…Nowadays, the RF algorithm is a promising method of machine learning, capable of solving the problems of image recognition, prediction and selection of informative features [30]. A random forest is created on the basis of decision trees and has the same set of hyperparameters, while in RF the process of searching for the root node and the separation of object nodes is performed randomly [31,32].…”

Section: Algorithm 2 Particle Swarm Optimizationmentioning

confidence: 99%

Development of Smart Technology for Complex Objects Prediction and Control on the Basis of a Distributed Control System and an Artificial Immune Systems Approach

Pschegusova¹,

Samigulina²

2019

Adv. sci. technol. eng. syst. j.

View full text Add to dashboard Cite

This paper is an extension of work originally presented in 2018 Global Smart Industry Conference (GloSIC). Researches are devoted to the development of Smart technology for complex objects control and prediction on the basis of a distributed Honeywell DCS control system of the TengizChevroil enterprise using the example of a technological process of medium pressure gas cleaning. The article describes how on the basis of the multialgorithm approach there was developed a modified algorithm based on modern artificial intelligence methods in order to select informative features (principal component method, Random Forest algorithm, particle swarm algorithm) and artificial immune systems (clonal selection) solving the image recognition problem and predicting the state of a complex control object. There was conducted a comparative analysis of the simulation results using the example of real production data (daily data of sensors from an average pressure absorber).

show abstract

“…However, the IA still has the disadvantages of a slow convergence rate in the late search stage and low accuracy. Therefore, it is necessary to make modifications to improve the performance of IA [26].…”

Section: Introductionmentioning

confidence: 99%

Spatial Optimization of Residential Care Facility Configuration Based on the Integration of Modified Immune Algorithm and GIS: A Case Study of Jing’an District in Shanghai, China

Cheng

Cui

2020

IJERPH

View full text Add to dashboard Cite

As the population is aging rapidly, the irrationality of residential care facility (RCF) configuration has impacted the efficiency and quality of the aged care services so significantly that the optimization of RCF configuration is urgently required. A multi-objective spatial optimization model for the RCF configuration is developed by considering the demands of three stakeholders, including the government, the elderly, and the investor. A modified immune algorithm (MIA) is implemented to find the optimal solutions, and the geographic information system (GIS) is used to extract information on spatial relationships and visually display optimization results. Jing’an District, part of Shanghai, China, is analyzed as a case study to demonstrate the advantages of this integrated approach. The configuration rationality of existing residential care facilities (RCFs) is analyzed, and a detailed recommendation for optimization is proposed. The results indicate that the number of existing RCFs is deficient; the locations of some RCFs are unreasonable, and there is a large gap between the service supply of existing RCFs and the demands of the elderly. To fully meet the care demands of the elderly, 6 new facilities containing 1193 beds are needed to be added. In comparison with the optimization results of other algorithms, MIA is superior in terms of the calculation accuracy and convergence rate. Based on the integration of MIA and GIS, the quantity, locations, and scale of RCFs can be optimized simultaneously, effectively, and comprehensively. The optimization scheme has improved the equity and efficiency of RCF configuration, increased the profits of investors, and reduced the travel costs of the elderly. The proposed method and optimization results have reference value for policy-making and planning of RCFs as well as other public service facilities.

show abstract

Modified immune network algorithm based on the Random Forest approach for the complex objects control

Cited by 24 publications

References 13 publications

Ontological model of multi-agent Smart-system for predicting drug properties based on modified algorithms of artificial immune systems

Ontological model of multi-agent Smart-system for predicting drug properties based on modified algorithms of artificial immune systems

Development of Smart Technology for Complex Objects Prediction and Control on the Basis of a Distributed Control System and an Artificial Immune Systems Approach

Spatial Optimization of Residential Care Facility Configuration Based on the Integration of Modified Immune Algorithm and GIS: A Case Study of Jing’an District in Shanghai, China

Contact Info

Product

Resources

About