※一部利用できない機能があります
| 1.
論文 |
論文
|
Nakayama, Kenji
概要:
金沢大学大学院自然科学研究科情報システム<br />金沢大学工学部<br />A simultaneous frequency- and time-domain approximation method for discrete-time
…
filters is proposed. In the method, transfer function coefficients are divided into two subsets, X//1 and X//2, which are employed for optimizing a time response and a frequency response, respectively. Frequency and time responses are optimized through the iterative Chebyshev approximation method and a method of solving linear equations, respectively. At the rth iteration step, the maximum frequency response error, which appeared at the (r minus 1)th step, is minimized, and X//2**(**r** minus **1**) becomes X//2**(**r**). X//1**(**r**) is obtained from linear equations including X//2**(**r**) as a constant. The frequency response at the rth step is evaluated using the above obtained X//1**(**r**) and X//2**(**r**). This means the optimum time response is always guaranteed in the frequency-response approximation procedure. A design example of a symmetrical impulse response shows the new approach is more efficient than conventional methods from the filter order reduction viewpoint.
続きを見る
|
||||
| 2.
論文 |
論文
|
Hara, Kazuyuki ; Nakayama, Kenji
概要:
金沢大学大学院自然科学研究科知能情報・数理<br />A training data selection method is proposed for multilayer neural networks (MLNNs). This met
…
hod selects a small number of the training data, which guarantee both generalization and fast training of the MLNNs applied to pattern classification. The generalization will be satisfied using the data locate close to the boundary of the pattern classes. However, if these data are only used in the training, convergence is slow. This phenomenon is analyzed in this paper. Therefore, in the proposed method, the MLNN is first trained using some number of the data, which are randomly selected (Step 1). The data, for which the output error is relatively large, are selected. Furthermore, they are paired with the nearest data belong to the different class. The newly selected data are further paired with the nearest data. Finally, pairs of the data, which locate close to the boundary, can be found. Using these pairs of the data, the MLNNs are further trained (Step 2). Since, there are some variations to combine Steps 1 and 2, the proposed method can be applied to both off-line and on-line training. The proposed method can reduce the number of the training data, at the same time, can hasten the training. Usefulness is confirmed through computer simulation.
続きを見る
|
||||
| 3.
論文 |
論文
|
Nakayama, Kenji ; Kato, Takuo ; Katayama, Hiroshi
|
||||
| 4.
論文 |
論文
|
Nakayama, Kenji
概要:
A bandlimited signal extrapolation algorithm is proposed. J. A. Cadzow's algorithm (1979) is modified to eliminate undesired outband spectra. An inverse filter stopband response is relaxed by adding small random numbers. A
…
constrained heuristic optimization is suggested that can use arbitrary signal properties as constraints. Through numerical examples, it is shown how regularization techniques can improve SNR (signal-to-noise-ratio) by 20 dB.
続きを見る
|
||||
| 5.
論文 |
論文
|
Keeni, Kanad ; Shimodaira, Hiroshi ; Nakayama, Kenji
概要:
This paper presents an automatic coding scheme for representing the output layer of a neural network. Compared to local representation where the number of output unit is p, the number of output unit required for the proposed representation is
…
close to log p. The output of seven different printers were used for evaluating the performance of the system. The proposed automatic representation gave the average recognition rate of 98.7% for 71 categories.
続きを見る
|
||||
| 6.
論文 |
論文
|
Wang, Youhua ; Ikeda, Kazushi ; Nakayama, Kenji
|
||||
| 7.
論文 |
論文
|
Nakayama, Kenji ; Katayama, Hiroshi
|
||||
| 8.
論文 |
論文
|
Ikeda, Kazushi ; Suzuki, Akihiro ; Nakayama, Kenji
概要:
The effects of the quantization of the parameters of a learning machine are discussed. The learning coefficient should be as small as possible for a better estimate of parameters. On the other hand, when the parameters are quantized, it
…
should be relatively larger in order to avoid the paralysis of learning originated from the quantization. How to choose the learning coefficient is given in this paper from the statistical point of view.
続きを見る
|
||||
| 9.
論文 |
論文
|
Nakayama, Kenji ; Imai, Kunihiko
概要:
金沢大学理工研究域電子情報学系<br />A neural demodulator is proposed for amplitude shift keying (ASK) signal. It has several important
…
features compared with conventional linear methods. First, necessary functions for ASK demodulation, including wide-band noise rejection, pule waveform shaping, and decoding, can be embodied in a single neural network. This means these functions are not separately designed but unified in a learning and organizing process. Second, these functions can be self-organized through the learning. Supervised learning algorithms, such as the backpropagation algorithm, can be applied for this purpose. Finally, both wide-band noise rejection and a very sharp waveform response can be simultaneously achieved. It is very difficult to be done by linear filtering. Computer simulation demonstrates efficiency of the proposed method.
続きを見る
|
||||
| 10.
論文 |
論文
|
Miyoshi, Seiji ; Nakayama, Kenji
概要:
In this paper, the geometric learning algorithm (GLA) is proposed for an elementary perceptron which includes a single output neuron. The GLA is a modified version of the affine projection algorithm (APA) for adaptive filters.
…
The weights update vector is determined geometrically towards the intersection of the k hyperplanes which are perpendicular to patterns to be classified. k is the order of the GLA. In the case of the APA, the target of the coefficients update is a single point which corresponds to the best identification of the unknown system. On the other hand, in the case of the GLA, the target of the weights update is an area, in which all the given patterns are classified correctly. Thus, their convergence conditions are different. In this paper, the convergence condition of the 1st order GLA for 2 patterns is theoretically derived. The new concept `the angle of the solution area' is introduced. The computer simulation results support that this new concept is a good estimation of the convergence properties.
続きを見る
|
