输入层自构造神经网络及其在光谱分析中的应用

为了解决红外光谱定量分析中的特征提取和校正规模问题，提出了一种输入层自构造神经网络。这种网络能够利用训练数据的某些先验知识，自然选择输入层神经元的个数。在学习过程中，输入神经元个数从最小值1开始，根据网络误差的变化逐步增加，最终确定最佳神经元数量。这种网络模型将特征提取和参数学习过程融一体，有利于提高建模效率。利用仿真红外光谱的定量分析实验表明，这种网络模型不仅能够对光谱数据实现高效率的波长选择，并具有抑制随机噪声和非线性干扰的能力。     

输入层自构造神经网络及其在光谱分析中的应用

为了解决红外光谱定量分析中的特征提取和校正规模问题,提出了一种输入层自构造神经网络。这种网络能够利用训练数据的某些先验知识,自动选择输入层神经元的个数。在学习过程中,输入神经元个数从最小值1开始,根据网络误差的变化逐步增加,最终确定最佳神经元数量。这种网络模型将特征提取和参数学习过程融为一体,有利于提高建模效率。利用仿真红外光谱的定量分析实验表明,这种网络模型不仅能够对光谱数据实现高效率的波长选择,并具有抑制随机噪声和非线性干扰的能力。     

BiGRU-Att-KF话题热度预测

为提高话题热度预测精度，提出一种面向话题热度趋势变化的BiGRU-Att-KF预测方法。选取微博文本总数作为话题热度的量化指标，构建话题热度时间序列；采用RMSE确定BiGRU的超参数神经元个数和批处理大小；利用BiGRU挖掘话题热度时间序列的全局特征信息，引入注意力机制自适应提取特征信息；采用卡尔曼滤波对注意力层输出结果进行动态调整。实验结果表明，BiGRU-Att-KF相较BiGRU-Att等方法，具有更优的鲁棒性和预测精度，能够更好模拟话题热度的变化趋势。     

小脑模型CMAC网络结构及有关参数的确定

从映射的角度分析了CMAC模型各层神经元之间的关系，根据网络输入向量的量化级数、泛化参数、相邻量化级引起的重叠神经元的个数，从理论上给出了虚拟层神经元数目的范围。对于存在实际层神经元的CMAC模型，讨论了压缩映射对网络学习收敛性的影响。最后通过机器手逆运动学问题的仿真实验，进一步比较说明了在从虚拟层神经元到实际层神经元的压缩映射中不同压缩比对网络学习收敛性及系统运行精度的影响。     

基于神经网络的自动售货机销量预测

采用神经网络理论对罐装饮料自动售货机销售额的预测进行了数学建模,采用BP神经网络实现了罐装饮料自动售货机销售额的预测,并通过隐层神经元和输入层神经元个数的确定以及训练算法性能对比验证了所设计模型和方法的可行性和有效性,为罐装饮料自动售货机的管理工作提供了一种新的方法.     

基于决策树算法确定一个最小神经网络

神经网络隐层神经元的个数对于网络的性能有着重要的影响,通常情况下,对于一个特定问题来说,没有一个确定的方法来决定隐含层到底应该有多少个神经元,一般采用试探的方法通过多次实验来达到理想效果.在分类问题中,决策树和神经网络的结构有着一定的关联性,通过把决策树映射到神经网络结构中来确定隐层神经元的个数的方法能够有效地设计神经网络的结构,从而提高训练的效率并达到良好的分类效果.实验结果表明,该方法能够得到一个有着良好识别率的最小神经网络.方法简单有效,直观且易于操作.     

一种改进的RBF神经网络DOA估计方法

提出一种优化径向基函数神经网络来波方位(DOA)估计模型结构和参数的方法。利用误差准则函数的收敛性,合理确定模型的隐层神经元数目,根据阵列信号相位差特征的空间分布特点,选择具有代表性的隐层神经元的中心,构建的RBF神经网络更能反映阵列的测向能力。相比于目前的径向基函数神经网络测向模型的构建方法,改进的DOA估计模型具有更好的泛化性能,能够提高测向精度。实验结果验证了该方法的有效性。     

一种新的多层感知机隐含层神经元个数上限计算方法

多层感知机在分类问题中具有广泛的应用。本文针对超平面阈值神经元构成的多层感知机用于分类的情况,求出了输入层神经元最多能把输入空间划分的区域数的解析表达式。该指标在很大程度上说明了感知机输入层的分类能力。本文还对隐含层神经元个数和输入层神经元个数之间的约束关系进行了讨论,得到了更准确的隐含层神经元个数上
上限。当分类空间的雏数远小于输入层神经元个数时,本文得到的隐含层神经元个数上限比现有的结果更小。     

前向神经网络合理隐含层结点个数估计

合理选择隐含层神经元个数是前向神经网络构造中的一个关键问题,对网络的泛化能力、训练速度等都具有重要的影响。该文提出了基于隐含层神经元输出之间的相关分析而进行隐含层神经元合理个数的估计方法,首先建立了基于网络输出和基于网络输出对网络各输入一阶偏导数的隐含层各神经元输出之间的相关程度度量,进而给出了基于模糊等价关系分析的神经元合理个数估计方法。具体应用结果证明了所提出方法的有效性。     

基于BP网络曲线拟合方法的研究

在利用BP神经网络进行曲线拟合时,为了解决如何确定BP神经网络隐含层神经元数问题,提出了一种新的快速构建BP神经网络结构的方法,即如何由输入层神经元数、输出层神经元数及样本点数来确定隐含层神经元数,同时针对在曲线拟合过程中经常出现的一些问题提出了解决方案。实验结果表明,该构建方法和改进方案在提高曲线的拟合精度、加快收敛速度方面收到了较好的效果．     

AN ADAPTIVE EXPECTATION GENETIC ALGORITHM BASED ON ANFIS AND MULTINATIONAL STOCK MARKET VOLATILITY CAUSALITY FOR TAIEX FORECASTING

The stock market is a highly complex and dynamic system, and forecasting stock is complicated and difficult. Successful prediction of stock prices may promise attractive benefits; therefore, stock market forecasting is important and of great interest. The economy of Taiwan relies on international trade deeply and the fluctuations of international stock markets impact Taiwan's stock market to certain degree. It is practical to use the fluctuations of other stock markets as forecasting factors for forecasting on the Taiwan stock market. Further, stock market investors usually make short-term decisions based on recent price fluctuations, but most time series models use only the last period of stock price in forecasting. In this article, the proposed model uses the fluctuations of other national stock markets as forecasting factors and employs an expectation equation method whose parameters are optimized by a genetic algorithm (GA) joined with an adaptive network–based fuzzy inference system (ANFIS) model to forecast the Taiwan stock index. To evaluate the forecasting performance, the proposed model is compared with Chen's model and Yu's model. The experimental results indicate that the proposed model is superior to the listing methods (Chen's model and Yu's model) in terms of root mean squared error (RMSE).     

基于三次样条权函数神经网络的股价预测

随着经济的发展,股票投资已成为很多人的一种投资理财方式,而股票价格的预测也成为投资者关心和研究的焦点。建立一个运算速度和精确度都比较高的股价预测模型,对于金融投资者具有理论指导意义和实际应用价值。文中针对传统BP算法存在的学习速度慢、容易陷入局部极小值、隐层数不易确定等问题,使用三次样条权函数神经网络建立股价预测模型,克服了传统神经网络的缺点。仿真结果表明,该模型具有较高的预测精度,能够对股市进行有效的预测。     

基于生成式对抗神经网络的股票预测研究

针对股票价格具有非线性、非平稳的特点,提出一种结合自注意力机制和残差网络的生成式对抗神经网络模型（SAR-GAN）。该模型的生成器（generator）由长短期记忆网络（LSTM）层、自注意力机制层、残差层等构建而成,用于生成所预测股票的价格;判别器（discriminator）用于鉴别生成的股票价格与真实的股票价格。为验证模型良好的泛化性,选取上证指数及不同股票市场的热点行业龙头股票进行预测实验。实验结果表明,与LSTM、GRU、CNN-LSTM、CNN-GRU等模型相比,SAR-GAN模型能不同程度地减少预测误差。     

Support vector regression with chaos-based firefly algorithm for stock market price forecasting

Due to the inherent non-linearity and non-stationary characteristics of financial stock market price time series, conventional modeling techniques such as the Box–Jenkins autoregressive integrated moving average (ARIMA) are not adequate for stock market price forecasting. In this paper, a forecasting model based on chaotic mapping, firefly algorithm, and support vector regression (SVR) is proposed to predict stock market price. The forecasting model has three stages. In the first stage, a delay coordinate embedding method is used to reconstruct unseen phase space dynamics. In the second stage, a chaotic firefly algorithm is employed to optimize SVR hyperparameters. Finally in the third stage, the optimized SVR is used to forecast stock market price. The significance of the proposed algorithm is 3-fold. First, it integrates both chaos theory and the firefly algorithm to optimize SVR hyperparameters, whereas previous studies employ a genetic algorithm (GA) to optimize these parameters. Second, it uses a delay coordinate embedding method to reconstruct phase space dynamics. Third, it has high prediction accuracy due to its implementation of structural risk minimization (SRM). To show the applicability and superiority of the proposed algorithm, we selected the three most challenging stock market time series data from NASDAQ historical quotes, namely Intel, National Bank shares and Microsoft daily closed (last) stock price, and applied the proposed algorithm to these data. Compared with genetic algorithm-based SVR (SVR-GA), chaotic genetic algorithm-based SVR (SVR-CGA), firefly-based SVR (SVR-FA), artificial neural networks (ANNs) and adaptive neuro-fuzzy inference systems (ANFIS), the proposed model performs best based on two error measures, namely mean squared error (MSE) and mean absolute percent error (MAPE).     

A neural network model for estimating option prices

A neural network model that processes financial input data is developed to estimate the market price of options at closing. The network's ability to estimate closing prices is compared to the Black-Scholes model, the most widely used model for the pricing of options. Comparisons reveal that the mean squared error for the neural network is less than that of the Black-Scholes model in about half of the cases examined. The differences and similarities in the two modeling approaches are discussed. The neural network, which uses the same financial data as the Black-Scholes model, requires no distribution assumptions and learns the relationships between the financial input data and the option price from the historical data. The option-valuation equilibrium model of Black-Scholes determines option prices under the assumptions that prices follow a continuous time path and that the instantaneous volatility is nonstochastic.     

Integrating metaheuristics and Artificial Neural Networks for improved stock price prediction

Stock market price is one of the most important indicators of a country's economic growth. That's why determining the exact movements of stock market price is considerably regarded. However, complex and uncertain behaviors of stock market make exact determination impossible and hence strong forecasting models are deeply desirable for investors' financial decision making process. This study aims at evaluating the effectiveness of using technical indicators, such as simple moving average of close price, momentum close price, etc. in Turkish stock market. To capture the relationship between the technical indicators and the stock market for the period under investigation, hybrid Artificial Neural Network (ANN) models, which consist in exploiting capabilities of Harmony Search (HS) and Genetic Algorithm (GA), are used for selecting the most relevant technical indicators. In addition, this study simultaneously searches the most appropriate number of hidden neurons in hidden layer and in this respect; proposed models mitigate well-known problem of overfitting/underfitting of ANN. The comparison for each proposed model is done in four viewpoints: loss functions, return from investment analysis, buy and hold analysis, and graphical analysis. According to the statistical and financial performance of these models, HS based ANN model is found as a dominant model for stock market forecasting.     

Multiagent System Simulations of Treasury Auctions

This study uses a multiagent system to determine which payment rule provides the most revenue in treasury auctions. The agents learn how to bid using straightforward bid adjustment rules that are based on impulse balance learning. The market model encompasses the when-issued, auction, and secondary markets, as well as bidding constraints for primary dealers. I find that when the number of primary bidders is less than 13 (Canada) the Discriminatory payment rule is revenue superior to the Uniform payment across most market price spreads. When the number of primary bidders is greater than 14 (United States), Uniform payment is revenue superior to Discriminatory payment for all market price spreads. In general, revenue increases with the minimum bid constraint and with the number of primary dealers for Uniform, Average, and Vickrey payment rules.      

基于LSTM的股票价格预测建模与分析

股价波动是一个高度复杂的非线性系统，其股票的调整不是按照均匀的时间过程推进，具有自身的推进过程。结合LSTM（Long Short-Term Memory）递归神经网络的特性和股票市场的特点，对数据进行插值、小波降噪、归一化等预处理操作后，推送到搭建的不同LSTM层数与相同层数下不同隐藏神经元个数的LSTM网络模型中进行训练与测试。对比评价指标与预测效果找到适宜的LSTM层数与隐藏神经元个数，提高了预测准确率约30%。测试结果表明，该模型计算复杂度小，预测准确率有所提高，不仅能在股票投资前对预测股票走势提供有益的参考，还能帮助投资者在对实际股价有了进一步的认知后构建合适的股票投资策略。     

Forecasting volatility of oil price using an artificial neural network-GARCH model

This paper builds on previous research and seeks to determine whether improvements can be achieved in the forecasting of oil price volatility by using a hybrid model and incorporating financial variables. The main conclusion is that the hybrid model increases the volatility forecasting precision by 30% over previous models as measured by a heteroscedasticity-adjusted mean squared error (HMSE) model. Key financial variables included in the model that improved the prediction are the Euro/Dollar and Yen/Dollar exchange rates, and the DJIA and FTSE stock market indexes.     

Fusion ANFIS models based on multi-stock volatility causality for TAIEX forecasting

Stock market investors value accurate forecasting of future stock price from trading systems because of the potential for large profits. Thus, investors use different forecasting models, such as the time-series model, to assemble a superior investment portfolio. Unfortunately, there are three major drawbacks to the time-series model: (1) most statistical methods rely on some assumptions about the variables; (2) most conventional time-series models use only one variable in forecasting; and (3) the rules mined from artificial neural networks are not easily understandable. To address these shortcomings, this study proposes a new model based on multi-stock volatility causality, a fusion adaptive-network-based fuzzy inference system (ANFIS) procedure, for forecasting stock price problems in Taiwan. Furthermore, to illustrate the proposed model, three practical, collected stock index datasets from the USA and Taiwan stock markets are used in the empirical experiment. The experimental results indicate that the proposed model is superior to the listing methods in terms of root mean squared error, and further evaluation reveals that the profits comparison results for the proposed model produce higher profits than the listing models.