Generalization Capabilities of Feedforward Neural Networks for Pattern Recognition

This paper studies the generalization capability of feedforward neural networks (FNN).The mechanism of FNNs for classification is investigated from the geometric and probabilistic viewpoints. It is pointed out that the outputs of the output layer in the FNNs for classification correspond to the estimates of posteriori probability of the input pattern samples with desired outputs 1 or 0. The theorem for the generalized kernel function in the radial basis function networks (RBFN) is given. For an 2-layer perceptron network (2-LPN). an idea of using extended samples to improve generalization capability is proposed. Finally. the experimental results of radar target classification are given to verify the generaliztion capability of the RBFNs.

An adaptive physics-informed deep learning method for pore pressure prediction using seismic data

Accurate prediction of formation pore pressure is essential to predict fluid flow and manage hydrocarbon production in petroleum engineering.Recent deep learning technique has been receiving more interest due to the great potential to deal with pore pressure prediction.However,most of the traditional deep learning models are less efficient to address generalization problems.To fill this technical gap,in this work,we developed a new adaptive physics-informed deep learning model with high generalization capability to predict pore pressure values directly from seismic data.Specifically,the new model,named CGP-NN,consists of a novel parametric features extraction approach(1DCPP),a stacked multilayer gated recurrent model(multilayer GRU),and an adaptive physics-informed loss function.Through machine training,the developed model can automatically select the optimal physical model to constrain the results for each pore pressure prediction.The CGP-NN model has the best generalization when the physicsrelated metricλ=0.5.A hybrid approach combining Eaton and Bowers methods is also proposed to build machine-learnable labels for solving the problem of few labels.To validate the developed model and methodology,a case study on a complex reservoir in Tarim Basin was further performed to demonstrate the high accuracy on the pore pressure prediction of new wells along with the strong generalization ability.The adaptive physics-informed deep learning approach presented here has potential application in the prediction of pore pressures coupled with multiple genesis mechanisms using seismic data.

Rolling force prediction for strip casting using theoretical model and artificial intelligence

Rolling force for strip casting of 1Cr17 ferritic stainless steel was predicted using theoretical model and artificial intelligence.Solution zone was classified into two parts by kiss point position during casting strip.Navier-Stokes equation in fluid mechanics and stream function were introduced to analyze the rheological property of liquid zone and mushy zone,and deduce the analytic equation of unit compression stress distribution.The traditional hot rolling model was still used in the solid zone.Neural networks based on feedforward training algorithm in Bayesian regularization were introduced to build model for kiss point position.The results show that calculation accuracy for verification data of 94.67% is in the range of ±7.0%,which indicates that the predicting accuracy of this model is very high.

Study on the Overfitting of the Artificial Neural Network Forecasting Model

Because of overfitting and the improvement of generalization capability (GC)available in the construction of forecasting models using artificial neural network (ANN), a newmethod is proposed for model establishment by means of making a low-dimension ANN learning matrixthrough principal component analysis (PCA). The results show that the PC A is able to construct anANN model without the need of finding an optimal structure with the appropriate number ofhidden-layer nodes, thus avoids overfitting by condensing forecasting information, reducingdimension and removing noise, and GC is greatly raised compared to the traditional ANN and stepwiseregression techniques for model establishment.

A novel overlapping minimization SMOTE algorithm for imbalanced classification

The synthetic minority oversampling technique(SMOTE) is a popular algorithm to reduce the impact of class imbalance in building classifiers, and has received several enhancements over the past 20 years. SMOTE and its variants synthesize a number of minority-class sample points in the original sample space to alleviate the adverse effects of class imbalance. This approach works well in many cases, but problems arise when synthetic sample points are generated in overlapping areas between different classes, which further complicates classifier training. To address this issue, this paper proposes a novel generalization-oriented rather than imputation-oriented minorityclass sample point generation algorithm, named overlapping minimization SMOTE(OM-SMOTE). This algorithm is designed specifically for binary imbalanced classification problems. OM-SMOTE first maps the original sample points into a new sample space by balancing sample encoding and classifier generalization. Then, OM-SMOTE employs a set of sophisticated minority-class sample point imputation rules to generate synthetic sample points that are as far as possible from overlapping areas between classes. Extensive experiments have been conducted on 32 imbalanced datasets to validate the effectiveness of OM-SMOTE. Results show that using OM-SMOTE to generate synthetic minority-class sample points leads to better classifier training performances for the naive Bayes,support vector machine, decision tree, and logistic regression classifiers than the 11 state-of-the-art SMOTE-based imputation algorithms. This demonstrates that OM-SMOTE is a viable approach for supporting the training of high-quality classifiers for imbalanced classification. The implementation of OM-SMOTE is shared publicly on the Git Hub platform at https://github.com/luxuan123123/OM-SMOTE/.