We present an alternative neural network architecture which is similar to the operation of a general fuzzy inference system. This hybrid fuzzy neural network (HFNN) is a modified multilayer feedforward neural network ...
详细信息
We present an alternative neural network architecture which is similar to the operation of a general fuzzy inference system. This hybrid fuzzy neural network (HFNN) is a modified multilayer feedforward neural network (MFNN) with four different layers. By using the gradient method, learning algorithms are derived. An example is presented to compare the approximation performance of the HFNN with the MFNN. The HFNN is then applied to an inverted pendulum control problem by using temporal backpropagation. The performance of the HFNN controller is illustrated by simulations.
This paper proposes a new double talk detection algorithm which is based on fuzzy theory for use in the adaptive echo canceller of a telecommunication system. In this method, the two inputs of the fuzzy inference for ...
详细信息
ISBN:
(纸本)0780329120
This paper proposes a new double talk detection algorithm which is based on fuzzy theory for use in the adaptive echo canceller of a telecommunication system. In this method, the two inputs of the fuzzy inference for detecting double talk condition are used. One input is the cross-correlation coefficient between the error signal and the primary signal which is the summed signal of the real echo signal and the near-end signal. The other is the cross-correlation coefficient between the estimation error signal and the primary signal. It is confirmed by computer simulation that this fuzzy double talk detector is able to track echo path variation accurately during the double talk.
The SAE 81C99 processor exhibits 4 different operation modes, 8 programmable fuzzy algorithms, and up to 256 inputs, 64 outputs, and 16,384 rules. The 1.0-mu m CMOS chip serves as a stand-alone device or as an on-chip...
详细信息
The SAE 81C99 processor exhibits 4 different operation modes, 8 programmable fuzzy algorithms, and up to 256 inputs, 64 outputs, and 16,384 rules. The 1.0-mu m CMOS chip serves as a stand-alone device or as an on-chip module for 8- or 16-bit microcontrollers. At 20-MHz crystal frequency and a maximum inference speed of 10 million rules/s, it supports very complex systems and millisecond (and faster) processes such as automotive electronics and pattern recognition.
This paper addresses the problem of identifying a probablistic FSM from its output. We propose a formal description of the FSM and its output, using a regular phrase-structured grammar. This description is associated ...
详细信息
This paper addresses the problem of identifying a probablistic FSM from its output. We propose a formal description of the FSM and its output, using a regular phrase-structured grammar. This description is associated with a cost measuring its information content. We then state the inference problem as a combinatorial optimization problem with an objective function define as the cost of the description of the FSM and its output. A heuristic algorithm is proposed which processes the output ''on line'' and yields a local minimum of our criterion. At each step (i.e., as new data is observed and processed), the procedure searches locally through the space of non-probabilistic FSMs, i.e., the transitions are initially regarded as being only present or absent. When a non-probabilistic model has been generated, the transition probabilities are determined from their relative frequencies given the behavior indicated by the data. This yields maximum likelihood estimates of the probabilities. The costs of the obtained probabilistic FSMs are then computed, and the K minimum cost FSMs are kept as starting points for the next step search. At each step, the generation of the FSMs is done via a local search through the neighborhoods of the K best FSMs obtained at the previous step. This algorithm is compared with similar algorithms proposed previously, and tested on a range of examples. It is found to work for a wider range of FSMs than the previous methods, and it is much more practical for large problems than previously proposed exhaustive search techniques.
The importance of resolving general queries in Bayesian networks has been the focus of attention in recent research on the Symbolic Probabilistic inference (SPI) algorithm [4], [5], SPI applies the concept of dependen...
详细信息
The importance of resolving general queries in Bayesian networks has been the focus of attention in recent research on the Symbolic Probabilistic inference (SPI) algorithm [4], [5], SPI applies the concept of dependency-directed backward search to probabilistic inference, and is incremental with respect to both queries and observations. Unlike traditional Bayesian network inferencing algorithms, the SPI algorithm is goal directed, performing only those calculations that are required to respond to queries. Research to date on SPI applies to Bayesian networks with only discrete-valued variables or only continuous variables (linear Gaussian) [3] and does not address networks with both discrete and continuous variables. In this paper, we extend the SPI algorithm to handle Bayesian networks made up of both discrete and continuous variables (SPI-DC). The only topological constraint of the networks is that the successors of any continuous variable have to be continuous variables as well. In order to have exact analytical solution, the relationships between the continuous variables are restricted to be ''linear Gaussian.'' With new representation, SPI-DC modifies the three basic SPI operations: multiplication, summation, and substitution. However, SPI-DC retains the framework of the SPI algorithm, namely building the search tree and recursive query mechanism and therefore retains the goal-directed and incrementality features of SPI.
In this paper, we explain the algorithms used in FINEST, the Fuzzy inference Environment Software with Tuning developed at LIFE (Laboratory for International Fuzzy Engineering Research). The research themes and associ...
详细信息
In this paper, we explain the algorithms used in FINEST, the Fuzzy inference Environment Software with Tuning developed at LIFE (Laboratory for International Fuzzy Engineering Research). The research themes and associated algorithms were defined to palliate the insufficiencies of usual inference methods and come naturally from the formulation of fuzzy "if... then..." rules. In particular, enhanced versions of combination operators, implication functions and aggregation operators are proposed, as well as a mechanism to tune the parameters used in the definition of the knowledge used in the system. Finally, one definition and formulation of backward reasoning with fuzzy "if... then..." rules is proposed.< >
In this paper, a theoretical framework for Bayesian adaptive training of the parameters of discrete hidden Markov model (DHMM) and of semi-continuous HMM (SCHMM) with Gaussian mixture state observation densities is pr...
详细信息
In this paper, a theoretical framework for Bayesian adaptive training of the parameters of discrete hidden Markov model (DHMM) and of semi-continuous HMM (SCHMM) with Gaussian mixture state observation densities is presented, In addition to formulating the forward-backward MAP (maximum a posteriori) and the segmental MAP algorithms for estimating the above HMM parameters, a computationally efficient segmental quasi-Bayes algorithm for estimating the state-specific mixture coefficients in SCHMM is developed, For estimating the parameters of the prior densities, a new empirical Bayes method based on the moment estimates is also proposed. The MAP algorithms and the prior parameter specification are directly applicable to training speaker adaptive HMM's, Practical issues related to the use of the proposed techniques for HMM-based speaker adaptation are studied, The proposed MAP algorithms are shown to be effective especially in the cases in which the training or adaptation data are limited.
We present a parallel architecture for fuzzy controllers and a methodology for their realization as analog CMOS chips for low- and medium-precision applications. These chips can be made to learn through the adaptation...
详细信息
We present a parallel architecture for fuzzy controllers and a methodology for their realization as analog CMOS chips for low- and medium-precision applications. These chips can be made to learn through the adaptation of electrically controllable parameters guided by a dedicated hardware-compatible learning algorithm. Our designs emphasize simplicity at the circuit level-a prerequisite for increasing processor complexity and operation speed. Examples include a three-input, four-rule controller chip in 1.5-mu m CMQS, single-poly, double-metal technology.
A new technique to recover the information loss in a block-based image coding system is developed in this paper, The proposed scheme is based on fuzzy logic reasoning and can be divided into three main steps: 1) hiera...
详细信息
A new technique to recover the information loss in a block-based image coding system is developed in this paper, The proposed scheme is based on fuzzy logic reasoning and can be divided into three main steps: 1) hierarchical compass interpolation/extrapolation (HCIE) in the spatial domain for initial recovery of lost blocks that mainly contain low-frequency information such as smooth background 2) coarse spectra interpretation by fuzzy logic reasoning for recovery of lost blocks that contain high-frequency information such as complex textures and fine features 3) sliding window iteration (SWI), which is performed in both spatial and spectral domains to efficiently integrate the results obtained in steps 1) and 2) such that the optimal result can be achieved in terms of surface continuity on block boundaries and a set of fuzzy inference rules. The proposed method, which is suitable for recovering both isolated and contiguous block losses, provides a new approach for error concealment of block-based image coding systems such as the JPEG coding standard and vector quantization-based coding algorithms, The principle of the proposed scheme can also be applied to block-based video compression schemes such as the H.261, MPEG, and HDTV standards, Simulation results are presented to illustrate the effectiveness of the proposed method,
暂无评论