In this paper, a FPGA implementation method for convolutional neural networks of deep learning is proposed. In order to solve the problem of efficiency and speed of depth learning algorithms, a Z convolutional neural network structure is presented. The structure is easy to implement with FPGA. At the same time, in order to meet the requirement of complex multi-input and multi-output practical application system and increase the speed of calculations, a 3 dimensional Z type convolutional neural network structure is proposed, and the function of hardware circuit is verified by circuit simulation.
This paper, as a first attempt, examines to design the recursive least-squares (RLS) finite impulse response (FIR) smoother, which estimates the signal at each start time of the finite interval in linear continuous-time stochastic systems. It is assumed that the signal is observed with additive white noise and is uncorrelated with the observation noise. It is a characteristic that the FIR smoother uses the covariance information of the signal process in the form of the semi-degenerate kernel and the variance of the observation noise besides the observed value. This paper also presents the recursive algorithm for the estimation error variance function of the RLS-FIR smoother to show the stability condition of the smoother.
This paper reports a pilot study on developing an instrument to predict the quality of e-commerce websites. The 8C model was adopted as the reference model of the heuristic evaluation. Each dimension of the 8C was mapped into a set of quantitative website elements, the websites were scraped to get the quantitative website elements, and the score of the dimension was calculated. A software was developed in PHP for the experiments. In the training process, 10 experiments were conducted and quantitative analysis was regressively conducted between the experiments. The conversion rate was used to verify the heuristic evaluation of an e-commerce website. The results showed that the mapping revisions between the experiments improved the performance of the evaluation instrument, therefore the experiment process and the quantitative mapping revision guideline proposed was on the right track. The experiment results and the future work have been discussed.
This paper tries to find out five poets’(Thomas Hardy, Wilde, Browning, Yeats, and Tagore) differences and similarities through analyzing their works on nineteenth Century by using natural language understanding technology and word vector model. Firstly, we collect enough poems from these five poets, build five corpus respectively, and calculate their high-frequency words, by using Natural Language Processing method. Then, based on the word vector model, we calculate the word vectors of the five poets’ high-frequency words, and combine the word vectors of each poet into one vector. Finally, we analyze the similarity between the combined word vectors by using the hierarchical clustering method. The result shows that the poems of Hardy, Browning, and Wilde are similar; the poems of Tagore and Yeats are relatively close—but the gap between the two is relatively large. In addition, we evaluate the stability of our approach by altering the word vector dimension, and try to analyze the results of clustering in a literary (poetic) perspective. Yeats and Tagore possessed a kind of mysticism poetics thought, while Hardy, Browning, and Wilde have the elements of realism combined with tragedy and comedy. The results are similar comparing to those we get from the word vector model.
This Research has critically examined the efficacy of a four layers Neural network with three inputs at each nodes and two hidden layers of three and four Neurons each before the output Neurons, to observe the pattern recognitions for a flood model. The non-linear neural network using an XOR inputs with 8- bits pattern has been used to predict the susceptible areas within the locations to flooding. Sequel to this, the research focused on the use of some control variables, namely; X1(Rate of soil consolidation), which can be classified as high ( low ( and stable ( The variable X2 with apparent resistivity range The X3 ranging and X4 with apparent resistivity range The algorithm that was used in training the neural network is the Back- propagation coded in c++ language with 300 epoch runs. The research however successfully classified and recognized the patterns for the areas susceptible to flooding.
This paper proposes radial basis function network (RBFN) models to estimate the head of gaseous petroleum fluids (GPFs) in electrical submersible pumps (ESPs) as an alternative to widely used empirical models. Both exact and efficient RBFN modelling approaches were employed. RBFN modelling essentially tend to minimise the modelling error, the discrepancy of estimated and real outputs within the modelling data. This may lead to overfitting and lack of model generality for operating conditions not reflected in modelling data. This critical matter was addressed in RBFN design process, and highly accurate RBFNs were developed and cross validated.