Results 1  10
of
19
Extraction of Rules from Discretetime Recurrent Neural Networks
, 1996
"... The extraction of symbolic knowledge from trained neural networks and the direct encoding of (partial) knowledge into networks prior to training are important issues. They allow the exchange of information between symbolic and connectionist knowledge representations. The focas of this paper is on t ..."
Abstract

Cited by 61 (15 self)
 Add to MetaCart
The extraction of symbolic knowledge from trained neural networks and the direct encoding of (partial) knowledge into networks prior to training are important issues. They allow the exchange of information between symbolic and connectionist knowledge representations. The focas of this paper is on the quality of the rules that are extracted from recurrent neural networks. Discretetime recurrent neural networks can be trained to correctly classify strings of a regular language. Rules defining the learned grammar can be extracted from networks in the form of deterministic finitestate automata (DFAs) by applying clustering algorithms in the output space of recurrent state neurons. Our algorithm can extract different finitestate automata that are consistent with a training set from the same network. We compare the generalization performances of these different models and the trained network and we introduce a heuristic that permits us to choose among the consistent DFAs the model which best approximates the learned regular grammar.
Noisy Time Series Prediction using a Recurrent Neural Network and Grammatical Inference
 Machine Learning
, 2001
"... Financial forecasting is an example of a signal processing problem which is challenging due to small sample sizes, high noise, nonstationarity, and nonlinearity. Neural networks have been very successful in a number of signal processing applications. We discuss fundamental limitations and inherent ..."
Abstract

Cited by 47 (0 self)
 Add to MetaCart
Financial forecasting is an example of a signal processing problem which is challenging due to small sample sizes, high noise, nonstationarity, and nonlinearity. Neural networks have been very successful in a number of signal processing applications. We discuss fundamental limitations and inherent difficulties when using neural networks for the processing of high noise, small sample size signals. We introduce a new intelligent signal processing method which addresses the difficulties. The method proposed uses conversion into a symbolic representation with a selforganizing map, and grammatical inference with recurrent neural networks. We apply the method to the prediction of daily foreign exchange rates, addressing difficulties with nonstationarity, overfitting, and unequal a priori class probabilities, and we find significant predictability in comprehensive experiments covering 5 different foreign exchange rates. The method correctly predicts the direction of change for th...
Symbolic Interpretation of Artificial Neural Networks
, 1996
"... Hybrid Intelligent Systems that combine knowledge based and artificial neural network systems typically have four phases involving domain knowledge representation, mapping of this knowledge into an initial connectionist architecture, network training and rule extraction respectively. The final phase ..."
Abstract

Cited by 33 (1 self)
 Add to MetaCart
Hybrid Intelligent Systems that combine knowledge based and artificial neural network systems typically have four phases involving domain knowledge representation, mapping of this knowledge into an initial connectionist architecture, network training and rule extraction respectively. The final phase is important because it can provide a trained connectionist architecture with explanation power and validate its output decisions. Moreover, it can be used to refine and maintain the initial knowledge acquired from domain experts. In this paper, we present three rule extraction techniques. The first technique extracts a set of binary rules from any type of neural network. The other two techniques are specific to feedforward networks with a single hidden layer of sigmoidal units. Technique 2 extracts partial rules that represent the most important embedded knowledge with an adjustable level of detail, while the third technique provides a more comprehensive and universal approach. A rule eval...
Computational Capabilities of Recurrent NARX Neural Networks
 IEEE Trans. on Systems, Man and Cybernetics
, 1997
"... Abstract—Recently, fully connected recurrent neural networks have been proven to be computationally rich—at least as powerful as Turing machines. This work focuses on another network which is popular in control applications and has been found to be very effective at learning a variety of problems. T ..."
Abstract

Cited by 31 (8 self)
 Add to MetaCart
Abstract—Recently, fully connected recurrent neural networks have been proven to be computationally rich—at least as powerful as Turing machines. This work focuses on another network which is popular in control applications and has been found to be very effective at learning a variety of problems. These networks are based upon Nonlinear AutoRegressive models with eXogenous Inputs (NARX models), and are therefore called NARX networks. As opposed to other recurrent networks, NARX networks have a limited feedback which comes only from the output neuron rather than from hidden states. They are formalized by y(t) =9(u(t0nu);111;u(t01); u(t);y(t0ny);111;y(t01)) where u(t) and y(t) represent input and output of the network at time t, nu and ny are the input and output order, and the function 9 is the mapping performed by a Multilayer Perceptron. We constructively prove that the NARX networks with a finite number of parameters are computationally as strong as fully connected recurrent networks and thus Turing machines. We conclude that in theory one can use the NARX models, rather than conventional recurrent networks without any computational loss even though their feedback is limited. Furthermore, these results raise the issue of what amount of feedback or recurrence is necessary for any network to be Turing equivalent and what restrictions on feedback limit computational power. I.
Rule Extraction from Recurrent Neural Networks: a Taxonomy and Review
 Neural Computation
, 2005
"... this paper, the progress of this development is reviewed and analysed in detail. In order to structure the survey and to evaluate the techniques, a taxonomy, specifically designed for this purpose, has been developed. Moreover, important open research issues are identified, that, if addressed pr ..."
Abstract

Cited by 24 (3 self)
 Add to MetaCart
this paper, the progress of this development is reviewed and analysed in detail. In order to structure the survey and to evaluate the techniques, a taxonomy, specifically designed for this purpose, has been developed. Moreover, important open research issues are identified, that, if addressed properly, possibly can give the field a significant push forward
Grammar Inference, Automata Induction, and Language Acquisition
 Handbook of Natural Language Processing
, 2000
"... The natural language learning problem has attracted the attention of researchers for several decades. Computational and formal models of language acquisition have provided some preliminary, yet promising insights of how children learn the language of their community. Further, these formal models als ..."
Abstract

Cited by 22 (3 self)
 Add to MetaCart
The natural language learning problem has attracted the attention of researchers for several decades. Computational and formal models of language acquisition have provided some preliminary, yet promising insights of how children learn the language of their community. Further, these formal models also provide an operational framework for the numerous practical applications of language learning. We will survey some of the key results in formal language learning. In particular, we will discuss the prominent computational approaches for learning different classes of formal languages and discuss how these fit in the broad context of natural language learning.
Recurrent Neural Networks With Small Weights Implement Definite Memory Machines
 NEURAL COMPUTATION
, 2003
"... Recent experimental studies indicate that recurrent neural networks initialized with `small' weights are inherently biased towards definite memory machines (Tino, Cernansky, Benuskova, 2002a; Tino, Cernansky, Benuskova, 2002b). This paper establishes a theoretical counterpart: transition funct ..."
Abstract

Cited by 21 (5 self)
 Add to MetaCart
Recent experimental studies indicate that recurrent neural networks initialized with `small' weights are inherently biased towards definite memory machines (Tino, Cernansky, Benuskova, 2002a; Tino, Cernansky, Benuskova, 2002b). This paper establishes a theoretical counterpart: transition function of recurrent network with small weights and `squashing ' activation function is a contraction. We prove that recurrent networks with contractive transition function can be approximated arbitrarily well on input sequences of unbounded length by a definite mem
The Neural Network Pushdown Automaton: Model, Stack and Learning Simulations
, 1993
"... In order for neural networks to learn complex languages or grammars, they must have sufficient computational power or resources to recognize or generate such languages. Though many approaches to effectively utilizing the computational power of neural networks have been discussed, an obvious one is t ..."
Abstract

Cited by 17 (2 self)
 Add to MetaCart
In order for neural networks to learn complex languages or grammars, they must have sufficient computational power or resources to recognize or generate such languages. Though many approaches to effectively utilizing the computational power of neural networks have been discussed, an obvious one is to couple a recurrent neural network with an external stack memory in effect creating a neural network pushdown automata (NNPDA). This NNPDA generalizes the concept of a recurrent network so that the network becomes a more complex computing structure. This paper discusses in detail a NNPDA its construction, how it can be trained and how useful symbolic information can be extracted from the trained network. To effectively couple the external stack to the neural network, an optimization method is developed which uses an error function that connects the learning of the state automaton of the neural network to the learning of the operation of the external stack: push, pop, and nooperation. To minimize the error function using gradient descent learning, an analog stack is designed such that the action and storage of information in the stack are continuous. One interpretation of a continuous stack is the probabilistic storage of and action on data. After training on sample strings of an unknown source grammar, a quantization procedure extracts from the analog stack and neural network a discrete pushdown automata (PDA). Simulations show that in learning deterministic contextfree grammars the balanced parenthesis language, 1 n 0 n, and the deterministic Palindrome the extracted PDA is correct in the sense that it can correctly recognize unseen strings of arbitrary length. In addition, the extracted PDAs can be shown to be identical or equivalent to the PDAs of the source grammars which were used to generate the training strings.
TimeDelay Neural Networks: Representation and Induction of Finite State Machines
 IEEE Transactions on Neural Networks
, 1997
"... In this work, we characterize and contrast the capabilities of the general class of timedelay neural networks (TDNN), with input delay neural networks (IDNN), the subclass of TDNNs with delays limited to the inputs. Each class of networks is capable of representing the same set of languages, those ..."
Abstract

Cited by 15 (6 self)
 Add to MetaCart
In this work, we characterize and contrast the capabilities of the general class of timedelay neural networks (TDNN), with input delay neural networks (IDNN), the subclass of TDNNs with delays limited to the inputs. Each class of networks is capable of representing the same set of languages, those embodied by the definite memory machines (DMM), a subclass of finite state machines. We demonstrate the close affinity between TDNNs and DMM languages by learning a very large DMM (2048 states) using only a few training examples. Even though both architectures are capable of representing the same class of languages, they have distinguishable learning biases. Intuition suggests that general TDNNs which include delays in hidden layers should perform well, compared to IDNNs, on problems in which the output can be expressed as a function on narrow input windows which repeat in time. On the other hand, these general TDNNs should perform poorly when the input windows are wide, or there is little r...
Natural Language Grammatical Inference: A Comparison of Recurrent Neural Networks and Machine Learning Methods
 Symbolic, Connectionist, and Statistical Approaches to Learning for Natural Language Processing, Lecture notes in AI
, 1996
"... We consider the task of training a neural network to classify natural language sentences as grammatical or ungrammatical, thereby exhibiting the same kind of discriminatory power provided by the Principles and Parameters linguistic framework, or Government and Binding theory. We investigate the foll ..."
Abstract

Cited by 13 (2 self)
 Add to MetaCart
We consider the task of training a neural network to classify natural language sentences as grammatical or ungrammatical, thereby exhibiting the same kind of discriminatory power provided by the Principles and Parameters linguistic framework, or Government and Binding theory. We investigate the following models: feedforward neural networks, FrasconiGoriSoda and BackTsoi locally recurrent neural networks, Williams and Zipser and Elman recurrent neural networks, Euclidean and editdistance nearestneighbors, and decision trees. Nonneural network machine learning methods are included primarily for comparison. We find that the Elman and Williams & Zipser recurrent neural networks are able to find a representation for the grammar which we believe is more parsimonious. These models exhibit the best performance. 1 Motivation 1.1 Representational Power of Recurrent Neural Networks Natural language has traditionally been handled using symbolic computation and recursive processes. The most ...