Abstract
In this study, artificial neural networks (ANNs) were used to reveal a quantitative relationship between catalytic composition and catalytic activity. This relationship was predefined using a hypothetical experimental space described by a multidimensional polynomial. The predictive ability of ANNs was investigated, i.e. an attempt was done to evaluate how ANNs can envisage a given hypothetical experimental space. Data sets for training, validation and testing of ANNs were obtained from the hypothetical experimental space using two different ways of sampling. Data were selected, (i) by means of our optimization algorithm called Holographic Research Strategy (HRS); and (ii) randomly. In order to model real experimentation, data were also generated with error. The relationship between the complexity of different network topologies and their predictive ability was investigated. It was shown that when data used for training have been perturbed with a given level of noise, less complex network architectures give acceptable accuracy. Additionally, estimated experimental spaces were visualized in a 2D layout by means of Holographic Mappings (HMs). Analysis of HMs revealed that ANNs trained by data sets obtained upon an optimization procedure provides better description of the experimental space in the vicinity of the optimum than ANNs trained by randomly selected data sets. This fact indicates again the importance of the optimization in combinatorial catalyst library design.
Keywords: Artificial neural networks (ANNs), high-throughput methods, information mining, knowledge extraction, library optimization, multidimensional space, visualization
Combinatorial Chemistry & High Throughput Screening
Title: Assessment of Predictive Ability of Artificial Neural Networks Using Holographic Mapping
Volume: 10 Issue: 2
Author(s): Andras Tompos, Lajos Vegvari, Erno Tfirst and Jozsef L. Margitfalvi
Affiliation:
Keywords: Artificial neural networks (ANNs), high-throughput methods, information mining, knowledge extraction, library optimization, multidimensional space, visualization
Abstract: In this study, artificial neural networks (ANNs) were used to reveal a quantitative relationship between catalytic composition and catalytic activity. This relationship was predefined using a hypothetical experimental space described by a multidimensional polynomial. The predictive ability of ANNs was investigated, i.e. an attempt was done to evaluate how ANNs can envisage a given hypothetical experimental space. Data sets for training, validation and testing of ANNs were obtained from the hypothetical experimental space using two different ways of sampling. Data were selected, (i) by means of our optimization algorithm called Holographic Research Strategy (HRS); and (ii) randomly. In order to model real experimentation, data were also generated with error. The relationship between the complexity of different network topologies and their predictive ability was investigated. It was shown that when data used for training have been perturbed with a given level of noise, less complex network architectures give acceptable accuracy. Additionally, estimated experimental spaces were visualized in a 2D layout by means of Holographic Mappings (HMs). Analysis of HMs revealed that ANNs trained by data sets obtained upon an optimization procedure provides better description of the experimental space in the vicinity of the optimum than ANNs trained by randomly selected data sets. This fact indicates again the importance of the optimization in combinatorial catalyst library design.
Export Options
About this article
Cite this article as:
Tompos Andras, Vegvari Lajos, Tfirst Erno and Margitfalvi L. Jozsef, Assessment of Predictive Ability of Artificial Neural Networks Using Holographic Mapping, Combinatorial Chemistry & High Throughput Screening 2007; 10 (2) . https://dx.doi.org/10.2174/138620707779940929
DOI https://dx.doi.org/10.2174/138620707779940929 |
Print ISSN 1386-2073 |
Publisher Name Bentham Science Publisher |
Online ISSN 1875-5402 |
Call for Papers in Thematic Issues
Artificial Intelligence Methods for Biomedical, Biochemical and Bioinformatics Problems
Recently, a large number of technologies based on artificial intelligence have been developed and applied to solve a diverse range of problems in the areas of biomedical, biochemical and bioinformatics problems. By utilizing powerful computing resources and massive amounts of data, methods based on artificial intelligence can significantly improve the ...read more
Eco-friendly Agents for Biological Control of Pathogenic Diseases
The discovery of an alternative biological approach to disease management includes work on medicinal products derived from natural sources as a starting point for the development of eco-friendly agents for these diseases and the injuries they cause, as well as reducing human contact with hazardous chemicals and their residues. We ...read more
Emerging trends in diseases mechanisms, noble drug targets and therapeutic strategies: focus on immunological and inflammatory disorders
Recently infectious and inflammatory diseases have been a key concern worldwide due to tremendous morbidity and mortality world Wide. Recent, nCOVID-9 pandemic is a good example for the emerging infectious disease outbreak. The world is facing many emerging and re-emerging diseases out breaks at present however, there is huge lack ...read more
Exploring Spectral Graph Theory in Combinatorial Chemistry
Scope of the Thematic Issue: Combinatorial chemistry involves the synthesis and analysis of a large number of diverse compounds simultaneously. Traditional methods rely on brute force experimentation, which can be time-consuming and resource-intensive. Spectral Graph Theory, a branch of mathematics dealing with the properties of graphs in relation to the ...read more
- Author Guidelines
- Graphical Abstracts
- Fabricating and Stating False Information
- Research Misconduct
- Post Publication Discussions and Corrections
- Publishing Ethics and Rectitude
- Increase Visibility of Your Article
- Archiving Policies
- Peer Review Workflow
- Order Your Article Before Print
- Promote Your Article
- Manuscript Transfer Facility
- Editorial Policies
- Allegations from Whistleblowers
Related Articles
-
Natural Compounds as Potential Anti-COVID Agents
Anti-Infective Agents SAR and Anti-Mycobacterial Activity of Quinolones and Triazoloquinolones: An Update
Anti-Infective Agents Electrochemical Nanomaterial-based Sensors/Biosensors for Drug Monitoring
Current Topics in Medicinal Chemistry Evidence for Complex Binding Profiles and Species Differences at the Translocator Protein (TSPO) (18 kDa)
Current Molecular Medicine The PTEN/PI3K/AKT Signalling Pathway in Cancer, Therapeutic Implications
Current Cancer Drug Targets Design and Synthesis of Thiazolidinedione Derivates as Selective PPARγ Modulators
Letters in Drug Design & Discovery How to Recognize and Workaround Pitfalls in QSAR Studies: A Critical Review
Current Medicinal Chemistry Exploitation of Some Natural Products for the Prevention and/or Nutritional Treatment of SARS-CoV2 Infection
Endocrine, Metabolic & Immune Disorders - Drug Targets Human β-secretase (BACE) and BACE Inhibitors: Progress Report
Current Topics in Medicinal Chemistry Recent Progress in Medicinal Chemistry of D4 Agonists
Current Medicinal Chemistry Development and Production of Commercial Therapeutic Monoclonal Antibodies in Mammalian Cell Expression Systems: An Overview of the Current Upstream Technologies
Current Pharmaceutical Biotechnology The Effect of Leverage and/or Influential on Structure-Activity Relationships
Combinatorial Chemistry & High Throughput Screening PPAR Ligands for Metabolic Disorders
Current Topics in Medicinal Chemistry On Electromagnetic Ablation of Biological Tissues: A Review
Recent Patents on Electrical Engineering Solution-Phase Parallel Synthesis of an Isoflavone Library for the Discovery of Novel Antigiardial Agents
Combinatorial Chemistry & High Throughput Screening The Analysis of SAR Position Error Effect on the Deceptive Jamming
International Journal of Sensors, Wireless Communications and Control High Mutation Rate Leads to Fitness Loss for Coronavirus Quasispecies
Coronaviruses Preface
Current Medicinal Chemistry Do Insecticides have Adverse Effects on Reproduction
Immunology, Endocrine & Metabolic Agents in Medicinal Chemistry (Discontinued) Protective Antiviral Antibodies that Lack Neutralizing Activity: Precedents and Evolution of Concepts
Current HIV Research