Results 1  10
of
53
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 58 (0 self)
 Add to MetaCart
(Show Context)
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...
Mining Scientific Data
, 2001
"... The past two decades have seen rapid advances in high performance computing and ..."
Abstract

Cited by 16 (4 self)
 Add to MetaCart
The past two decades have seen rapid advances in high performance computing and
Evaluation of air traffic complexity metrics using neural networks ans sector status
"... Abstract — This paper presents an original method to evaluate air traffic complexity metrics. Several complexity indicators, found in the litterature, were implemented and computed, using recorded radar data as input. A principal component analysis (PCA) provides some results on the correlations bet ..."
Abstract

Cited by 15 (6 self)
 Add to MetaCart
(Show Context)
Abstract — This paper presents an original method to evaluate air traffic complexity metrics. Several complexity indicators, found in the litterature, were implemented and computed, using recorded radar data as input. A principal component analysis (PCA) provides some results on the correlations between these indicators. Neural networks are then used to find a relationship between complexity indicators and the actual sector configurations. Assuming that the decisions to group or split sectors are somewhat related to the controllers workload, this method allows to identify which types of complexity indicators are significantly related to the actual workload. I.
Learning, Bayesian Probability, Graphical Models, and Abduction
 Abduction and Induction: Essays on their Relation and Integration, Chapter 10
, 1998
"... In this chapter I review Bayesian statistics as used for induction and relate it to logicbased abduction. Much reasoning under uncertainty, including induction, is based on Bayes' rule. Bayes' rule is interesting precisely because it provides a mechanism for abduction. I review work of Bu ..."
Abstract

Cited by 12 (0 self)
 Add to MetaCart
(Show Context)
In this chapter I review Bayesian statistics as used for induction and relate it to logicbased abduction. Much reasoning under uncertainty, including induction, is based on Bayes' rule. Bayes' rule is interesting precisely because it provides a mechanism for abduction. I review work of Buntine that argues that much of the work on Bayesian learning can be best viewed in terms of graphical models such as Bayesian networks, and review previous work of Poole that relates Bayesian networks to logicbased abduction. This lets us see how much of the work on induction can be viewed in terms of logicbased abduction. I then explore what this means for extending logicbased abduction to richer representations, such as learning decision trees with probabilities at the leaves. Much of this paper is tutorial in nature; both the probabilistic and logicbased notions of abduction and induction are introduced and motivated. 1 Introduction This paper explores the relationship between learning (induct...
Symbolic models and emergent models: A review
 IEEE Trans. Autonomous Mental Development
, 2012
"... Abstract—There exists a large conceptual gap between symbolic models and emergent models for the mind. Many emergent models work on lowlevel sensory data, while many symbolic models deal with highlevel abstract (i.e., action) symbols. There has been relatively little study on intermediate represen ..."
Abstract

Cited by 11 (7 self)
 Add to MetaCart
(Show Context)
Abstract—There exists a large conceptual gap between symbolic models and emergent models for the mind. Many emergent models work on lowlevel sensory data, while many symbolic models deal with highlevel abstract (i.e., action) symbols. There has been relatively little study on intermediate representations, mainly because of a lack of knowledge about how representations fully autonomously emerge inside the closed brain skull, using information from the exposed two ends (the sensory end and the motor end). As reviewed here, this situation is changing. A fundamental challenge for emergent modelsisabstraction,which symbolic models enjoy through human handcrafting. The term abstract refers to properties disassociated with any particular form. Emergent abstraction seems possible, although the brain appears to never receive a computer symbol (e.g., ASCII code) or produce such a symbol. This paper reviews major agent models with an emphasis on representation. It suggests two different ways to relate symbolic representations with emergent representations: One is based on their categorical definitions. The other considers that a symbolic representation corresponds to a brain’s outside behaviors observed and handcrafted by other outside human observers; but an emergent representation is inside the brain. Index Terms—Agents, attention, brain architecture, complexity, computer vision, emergent representation, graphic models, mental
Paper Why Have We Passed “Neural Networks Do Not Abstract Well”?
"... well. A Finite Automaton (FA) is a base net for many sophisticated probabilitybased systems of artificial intelligence, for statebased abstraction. However, an FA processes symbols, instead of images that the brain senses and produces (e.g., sensory images and effector images). This paper informal ..."
Abstract

Cited by 6 (5 self)
 Add to MetaCart
(Show Context)
well. A Finite Automaton (FA) is a base net for many sophisticated probabilitybased systems of artificial intelligence, for statebased abstraction. However, an FA processes symbols, instead of images that the brain senses and produces (e.g., sensory images and effector images). This paper informally introduces recent advances along the line of a new type of, brainanatomy inspired, neural networks —Developmental Networks (DNs). The new theoretical results discussed here include: (1) From any complex FA that demonstrates human knowledge through its sequence of the symbolic inputsoutputs, the Developmental Program (DP) of DN incrementally develops a corresponding DN through the image codes of the symbolic inputsoutputs of the FA. The DN learning from the FA is incremental, immediate and errorfree. (2) After learning the FA, if the DN freezes its learning but runs, it generalizes optimally for infinitely many image inputs and actions based on the embedded innerproduct distance, state equivalence, and the principle of maximum likelihood. (3) After learning the FA, if the DN continues to learn and run, it ―thinks‖ optimally in the sense of maximum likelihood based on its past experience. These three theoretical results have also been supported by experimental results using real images and text of natural languages. Together, they seem to argue that the neural networks as a class of methods has passed ―neural networks do not abstract well‖.
Soft Computing Methodologies for Structural Optimization
, 2003
"... The paper examines the efficiency of soft computing techniques in structural optimization, in particular algorithms based on evolution strategies combined with neural networks, for solving largescale, continuous or discrete structural optimization problems. The proposed combined algorithms are impl ..."
Abstract

Cited by 6 (0 self)
 Add to MetaCart
The paper examines the efficiency of soft computing techniques in structural optimization, in particular algorithms based on evolution strategies combined with neural networks, for solving largescale, continuous or discrete structural optimization problems. The proposed combined algorithms are implemented both in deterministic and reliability based structural optimization problems, in an effort to increase the computational efficiency as well as the robustness of the optimization procedure. The use of neural networks was motivated by the timeconsuming repeated finite element analyses required during the optimization process. A trained neural network is used to perform either the deterministic constraints check or, in the case of reliability based optimization, both the deterministic and the probabilistic constraints checks. The suitability of the neural network predictions is investigated in a number of structural optimization problems in order to demonstrate the computational advantages of the proposed methodologies.
Noisy time series prediction using symbolic representation and recurrent neural network grammatical inference
, 1996
"... 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 6 (0 self)
 Add to MetaCart
(Show Context)
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 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 the next day with an error rate of 47.1%. The error rate reduces to around 40% when rejecting examples where the system has low confidence in its prediction. The symbolic representation aids the extraction of symbolic knowledge from the recurrent neural networks in the form of deterministic finite state automata. These automata explain the operation of the system and are often relatively simple. Rules related to well known behavior such as trend following and mean reversal are extracted.
A Comparison between Single and Combined Backpropagation Neural Networks in the Prediction of Turnover.
 Applications of Artificial Intelligence
, 1998
"... Artificial neural networks are now being extensively used in the area of marketing analysis as they are well suited to this type of nonlinear problem. A retail company planned to improve its performance by using neural networks to predict turnover and data used in the experiment was provided by the ..."
Abstract

Cited by 6 (0 self)
 Add to MetaCart
(Show Context)
Artificial neural networks are now being extensively used in the area of marketing analysis as they are well suited to this type of nonlinear problem. A retail company planned to improve its performance by using neural networks to predict turnover and data used in the experiment was provided by the company. The study compares the performance of a combination of neural networks to that of a single neural network. The results show that backpropagation neural networks are effective tools which can give good results in solving a nonlinear prediction problem, even when data is poorly represented. 1 Introduction In this section we look at the problem area of forecasting and discuss issues concerning sales forecasting in business. We also give a brief introduction to artificial neural networks. Finally, the structure of this paper is outlined. 1.1 Problem area Forecasting the future has always been an exciting and interesting problem for researchers. Understanding the world and its event...
An efficient airspace configuration forecast
"... Abstract — This publication is the continuation of previous research which aims at improving the predictability and the flexibility of the airspace management process by computing realistic forecasts of the airspace configurations in Enroute ATC centers. In previous papers, we selected relevant com ..."
Abstract

Cited by 5 (2 self)
 Add to MetaCart
(Show Context)
Abstract — This publication is the continuation of previous research which aims at improving the predictability and the flexibility of the airspace management process by computing realistic forecasts of the airspace configurations in Enroute ATC centers. In previous papers, we selected relevant complexity metrics to predict the controllers workload, using neural networks trained on historical data. We also introduced new algorithms to build optimally balanced airspace configurations, exploring all possible combinations of elementary sectors. These workload prediction model and airspace partitionning algorithms were tested on real recorded traffic. In this paper, airspace configurations are forecast from planned traffic, using the CATS/OPAS simulator to compute trajectories from flight plans. The efficiency of the resulting airspace configurations is assessed by comparing to the actual FMP (Flow Management Position) prediction. Some preliminary developments of an experimental HMI that will be used to test and tune our algorithms are also presented.