Filters








28 Hits in 2.5 sec

Hybrid preference machines based on inspiration from neuroscience

Stefan Wermter, Christo Panchev
2002 Cognitive Systems Research  
However, very little research has been done on a framework which connects neuroscience-inspired models with connectionist models and higher level symbolic processing.  ...  It is a first hybrid framework which allows a link between spiking neural networks, connectionist preference machines and symbolic finite state machines.  ...  In particular we have exfor that column is not examined. plored the use of preference machines as one par- Table 2 shows the results obtained for the training ticular type of hybrid sequential machines  ... 
doi:10.1016/s1389-0417(01)00061-4 fatcat:u3xrqmrdirbztoyvkxi2k3iheu

Knowledge based modular networks for process modelling and control

J. Peres, R. Oliveira, S. Feyo de Azevedo
2001 Computers and Chemical Engineering  
The results show that it is possible to obtain more accurate process description when all available sources of knowledge are incorporated in the process model.  ...  This paper addresses methods of knowledge engineering for chemical and biochemical process modelling and control. The concept of Knowledge Based Modular (KBM) Networks is presented.  ...  Several authors have been proposing hybrid model structures for chemical and biochemical processes.  ... 
doi:10.1016/s0098-1354(01)00665-2 fatcat:xmxypses4vdqngkzsasoaj3kjq

Neuro-Fuzzy Paradigms for Intelligent Energy Management [chapter]

Ajith Abraham, Muhammad Riaz Khan
2004 Studies in Fuzziness and Soft Computing  
In the second approach, the input parameters consisting of load patterns and weather parameters are fuzzified and used to train a neural network.  ...  We applied the backpropagation algorithm to train neural network to find a preliminary forecast load.  ...  Experiments with all 3 data sets were repeated 3 times for all the connectionist models. • Neural network training We used a feedforward neural network with 2 hidden layers and trained using the backpropagation  ... 
doi:10.1007/978-3-540-39615-4_12 fatcat:iqa4egojlngbdbst2sys4y5kie

A Linear Genetic Programming Approach for Modelling Electricity Demand Prediction in Victoria [chapter]

Maumita Bhattacharya, Ajith Abraham, Baikunth Nath
2002 Hybrid Information Systems  
To evaluate, we considered load demand patterns for ten consecutive months taken every 30 minutes for training the different prediction models.  ...  Genetic programming (GP), a relatively young and growing branch of evolutionary computation is gradually proving to be a promising method of modelling complex prediction and classification problems.  ...  models can be minimized. iv) Logically GP does not require any formal architecture selection as required in many connectionist models.  ... 
doi:10.1007/978-3-7908-1782-9_28 fatcat:ixycwvvj2ngmljb45vrgofza44

Evolving fuzzy neural networks for supervised/unsupervised online knowledge-based learning

N. Kasabov
2001 IEEE Transactions on Systems Man and Cybernetics Part B (Cybernetics)  
Implementing NNMs of the ECOS framework require connectionist models that support these principles. Such model is the evolving fuzzy neural network (EFuNN). 3.  ...  EFuNNs evolve their structure and parameter values through incremental, hybrid supervised/unsupervised, on-line learning.  ...  I would like to thank he reviewers for their useful comments and suggestions that helped me to improve the paper significantly from its initial version.  ... 
doi:10.1109/3477.969494 pmid:18244856 fatcat:qi6jnbbmkneb3a7tj7oqezoqee

Machine Learning for Anomaly Detection: A Systematic Review

Ali Bou Nassif, Manar Abu Talib, Qassim Nasir, Fatima Mohamad Dakalbab
2021 IEEE Access  
Those ML techniques are used in two forms: standalone or hybrid models. Hybrid models are obtained by combining two or more ML techniques.  ...  Moreover, SVM is the most used technique as either standalone or in hybrid models.  ... 
doi:10.1109/access.2021.3083060 fatcat:vv7qthbvqjdz7ksm3yosulk22q

Extending the functional equivalence of radial basis function networks and fuzzy inference systems

K.J. Hunt, R. Haas, R. Murray-Smith
1996 IEEE Transactions on Neural Networks  
The more general framework allows the removal of some of the restrictive conditions of the previous result. the full Takagi-Sugeno model of fuzzy inference.  ...  The practical relevance of the functional equivalence result is that the leaning algorithm of one paradigm can be used to train models expressed in the other paradigm.  ...  and learning algorithms: and development of hybrid learning techniques mixing sym-While some aspects have already been used implicitly in numerous works on hybrid learning and approximation theoretic  ... 
doi:10.1109/72.501735 pmid:18263474 fatcat:rmp7nkohe5eorinivhhyiqi6ze

Genetics-Based Machine Learning [chapter]

Tim Kovacs
2012 Handbook of Natural Computing  
Touretsky, editor, Proc. 1990 Connectionist Models Summer School, pages 81-90. Morgan Kaufmann, 1990. 98 [70] Arjun Chandra and Xin Yao.  ...  Evolving networks: using the genetic algorithm with connectionistic learning. In C.G. Langton, C. Taylor, J.D. Farmer, and S.  ... 
doi:10.1007/978-3-540-92910-9_30 fatcat:rm5bx5lwdvfalolrky6lpyt67a

A Data-Driven Predictive Prognostic Model for Lithium-Ion Batteries based on a Deep Learning Algorithm

Phattara Khumprom, Nita Yodo
2019 Energies  
The main approach of PHM evaluation of the battery is to determine the State of Health (SoH) and the Remaining Useful Life (RUL) of the battery.  ...  The advancements of computational tools and big data algorithms have led to a new era of data-driven predictive analysis approaches, using machine learning algorithms.  ...  The modeling methods have grown from using only a single algorithm such as DNN, CNN, and RNN, to the Hybrid model, or a combination of multiple layer types and traditional algorithms.  ... 
doi:10.3390/en12040660 fatcat:mfgpqizagje7vecvmoflye3mxm

Variations of the two-spiral task

Stephan K. Chalup, Lukasz Wiklendt
2007 Connection science  
It was initially posted to the connectionist mailing list by Alexis P. Wieland and then included into the cmu Neural Network benchmark repository (White et al., 1995) .  ...  Geometrically the perceptron can be interpreted as a separating hyperplane and can be used for binary classification of linearly separable data.  ...  Acknowledgment The authors want to thank the anonymous reviewers for their useful comments.  ... 
doi:10.1080/09540090701398017 fatcat:uxgho6q2enbwfjebiffmjjbwky

Evolving connectionist systems: A theory and a case study on adaptive speech recognition

N. Kasabov
IJCNN'99. International Joint Conference on Neural Networks. Proceedings (Cat. No.99CH36339)  
The paper introduces evolving connectionist systems (ECOS) as an effective approach to building on-line, adaptive intelligent systems.  ...  ECOS evolve through incremental, hybrid (supervised/unsupervised), on-line learning. They can accommodate new input data, including new features, new classes, etc. through local element tuning.  ...  Implementing the ECOS framework and the NNM from it requires connectionist models that comply with the ECOS principles.  ... 
doi:10.1109/ijcnn.1999.836007 dblp:conf/ijcnn/Kasabov99 fatcat:lcsm27kx55bctf4x5lkxbtnpjy

C3W semantic Temporal Entanglement Modelling for Human - Machine Interfaces [chapter]

John Ronczka
2012 Semantics - Advances in Theories and Mathematical Models  
Secondly, WOSSI uses cognitive bases with hybrid semantics to facilitate moving from the macro, meso, micro, and quantum-nano scales with minimal support conjectures [9].  ...  Methodology The methodology focuses on benchmarking between how impairments in Humans, weapons systems, gaming (wargaming) systems are modeled using exercises.  ...  So what should it be F-semantics, I-semantics or a hybrid? F-semantics that is, the semantics is deterministic: no stable models or well-founded model is empty, but is meaningful.  ... 
doi:10.5772/38688 fatcat:neqc2335sjfizohlrvvqim2pdi

Multi-Domain Multi-Task Rehearsal for Lifelong Learning [article]

Fan Lyu, Shuai Wang, Wei Feng, Zihan Ye, Fuyuan Hu, Song Wang
2020 arXiv   pre-print
However, the old tasks of the most previous rehearsal-based methods suffer from the unpredictable domain shift when training the new task.  ...  Specifically, a two-level angular margin loss is proposed to encourage the intra-class/task compactness and inter-class/task discrepancy, which keeps the model from domain chaos.  ...  The previous methods update the model by the optimal gradient that highly rely on the angle between the gradients of old and new tasks. In contrast, we directly obtain the hybrid  ... 
arXiv:2012.07236v1 fatcat:2qqmpcoaffcuvizafrlmybseai

A Review on Machine Learning Approaches for Network Malicious Behavior Detection in Emerging Technologies

Mahdi Rabbani, Yongli Wang, Reza Khoshkangini, Hamed Jelodar, Ruxin Zhao, Sajjad Bagheri Baba Ahmadi, Seyedvalyallah Ayobi
2021 Entropy  
learning approaches including supervised, unsupervised, new deep and ensemble learning techniques have been comprehensively discussed; moreover, some details about currently available benchmark datasets for training  ...  [103] proposed a hybrid model for malicious detection using K-means, a back propagation neural network and naïve Bayes.  ...  [8] conducted a survey on IDSs using ensemble and hybrid learning systems.  ... 
doi:10.3390/e23050529 pmid:33923125 fatcat:d7sfiqhbkzhtre3vtl74eh67qy

A Robust Evolutionary Algorithm for Training Neural Networks

Jinn-Moon Yang, Cheng-Yan Kao
2001 Neural computing & applications (Print)  
A new evolutionary algorithm is introduced for training both feedforward and recurrent neural networks.  ...  All 2 N patterns are used in the training phase, and no validation set is used.  ...  Section 2 describes the model of artificial neural networks trained by our FCEA. Section 3 describes FCEA in detail, and gives motivations and ideas behind various design choices.  ... 
doi:10.1007/s521-001-8050-2 fatcat:3dhwcjtasvgtxcaho5ogbepk5q
« Previous Showing results 1 — 15 out of 28 results