Search

Kernel-based neural network (KNN) is proposed as a neuron that is applicable in online learning with adaptive parameters. This neuron with adaptive kernel parameter can classify data accurately instead of using a multilayer error back propagation neural network. The proposed method, whose heart is kernel least-mean-square, can reduce memory requirement with sparsification technique, and the kernel can adaptively spread. Our experiments will reveal that this method is much faster and more accurate than previous online learning algorithms. [ABSTRACT FROM AUTHOR]

This paper presents a new model of support vector machines ( SVMs) that handle data with tolerance and uncertainty. The constraints of the SVM are converted to fuzzy inequality. Giving more relaxation to the constraints allows us to consider an importance degree for each training samples in the constraints of the SVM. The new method is called relaxed constraints support vector machines ( RSVMs). Also, the fuzzy SVM model is improved with more relaxed constraints. The new model is called fuzzy RSVM. With this method, we are able to consider importance degree for training samples both in the cost function and constraints of the SVM, simultaneously. In addition, we extend our method to solve one-class classification problems. The effectiveness of the proposed method is demonstrated on artificial and real-life data sets. [ABSTRACT FROM AUTHOR]

This paper presents a human gait recognition algorithm based on a leg gesture separation. Main innovation in this paper is gait recognition using leg gesture classification which is invariant to covariate conditions during walking sequence and just focuses on underbody motions and a neuro-fuzzy combiner classifier (NFCC) which derives a high precision recognition system. At the end, performance of the proposed algorithm has been validated by using the HumanID Gait Challenge data set (HGCD), the largest gait benchmarking data set with 122 objects with different realistic parameters including viewpoint, shoe, surface, carrying condition, and time. And it has been compared to recent algorithm of gait recognition. [ABSTRACT FROM AUTHOR]

Support vector machines (SVM) is one of the prevalent techniques in the machine learning which is applicable in many of the real world classification problems. However, these problems are sensitive to noise and the presence of the outliers in training data. For this reason, other methods like fuzzy SVM (FSVM) were introduced to solve this challenge. Similar to SVM, these methods look for finding an optimal hyperplane that separates classes with the maximum possible margin. The main difference is the allocation of fuzzy membership degree to each training sample based on its significance so that leads to the resistance of the method. In this paper to overcome this problem, we proposed a robust FSVM based on prior knowledge related to error distribution (PKED-FSVM) to classify synthetic and real data. The optimal coefficient in our proposed PKED-FSVM method is introduced from the viewpoint of minimum Bayesian risk. Our proposed method is compared with SVM and other FSVM methods and results are obtained on several types of synthetic data from UCI. Finally, we collected the social media images from 500px based on user-tagged and created two data sets; Coastal-Non-Coastal, and the Rural–urban, and used the proposed method to classify these two data sets and separating coastal areas from non-coastal areas, and urban areas from rural areas, respectively. The results show that our proposed method has acceptable performance and works better than other competing methods. [ABSTRACT FROM AUTHOR]

• Deep multi-task networks have high uncertainty despite encouraging performance. • Multi-task learning system performance is sensitive to weighting strategy for tasks. • Paper proposes a novel weighting strategy to improve model uncertainty and accuracy. • This strategy enhances quality of weights by Bayesian estimator and particle filter. • Proposed tasks' weights make model have low uncertainty and high performance. Recently multi-task learning (MTL) has been widely used in different applications to build more robust models by sharing knowledge across several related tasks. However, one challenge that arises is the variability in the learning pace of different tasks causing the inefficiency of naively training all tasks. Therefore, it is of great importance to consider some coefficients to balance tasks in the process of learning, but, due to the large search space and the significance of setting them properly, conventional search methods such as grid or random search are no longer effective. In this paper, we propose a learning mechanism for these coefficients based on the high efficiency of the particle filter (PF) algorithm to deal with nonlinear search problems. PF considers each state of the tasks' coefficients as a particle and recursively converges coefficients to an optimum point. While in most previous works coefficients were evaluated to only increase performance, to address the recent concerns related to applying AI in real-world applications, we also incorporate uncertainty alongside our method to prevent learning coefficients leading to unstable outcomes. This mechanism is independent of the models main learning process and can be easily added to every learning system without changing its training algorithm. Extensive experiments on real-world data sets demonstrate the superiority of the proposed method over the state-of-the-art methods on both performance and uncertainty. We also proved the acceptable performance of the method using Cramer Rao lower bound theory. [ABSTRACT FROM AUTHOR]

• Mixture noise reduction as a key pre-process improves the HSI images' quality. • Mixture noise modeling is a challenge and should be close to the real noise. • Proposed mixture noise model leads to Bayesian framework selection. • Proposed method used low-rank factorization, total variation, L1, and nuclear norm. • Devised problem is solved by augmented Lagrange multiplier and proximal gradient. Mixture noise reduction is a key pre-process that improves the hyperspectral images' (HSI) quality and prepares them for the next processes. Mixture noise should be modeled as similar as possible to the real HSI noise model, which is a challenge for noise removal methods. These use noise model simplification assumptions that can take the method performance away from real scenarios. This paper adopts a new general model for the mixture noise which leads to one model selection framework. The optimization problem of the proposed method is formed using Bayesian risk in five different models and promoted to total variation regularized low-rank matrix factorization. The devised optimization problem is solved using augmented Lagrange multiplier and proximal gradient algorithms. Also, we compared the proposed method with other state-of-the-art methods to reduce the mixture noise of HSI, which is a combination of Gaussian (i.i.d or non-i.i.d), and sparse (e.g., stripe, deadline, impulse) noises. The results obtained on both real and synthetic HSI data sets show the proposed method's performance superiority to other competing methods in both visual comparisons and quantitative evaluations. [ABSTRACT FROM AUTHOR]

The conventional Linear Discriminant Analysis (LDA) model has some challenges, such as sensitivity to the outlier, the singularity problem of the within-class scatter matrix, and Gaussian assumption of data within the same class. This paper proposes a robust LDA method that tries to solve the sensitivity to outliers and singularity problems. Specifically, we first use Bayesian risk to design the proposed method optimization problem. Then, the proposed Density-oriented LDA (DLDA) method used the data density as prior knowledge for robustness against outliers. The proposed method can classify non-linear and multi-mode distribution data sets. Furthermore, the proposed method can be employed for big data classification using the AdaBoost approach. Experimental results on synthetic and real data sets demonstrate the proposed DLDA method's superiority over other competing methods. • LDA model has some challenges, e.g., sensitivity to outlier and singularity problem. • Using the data density as prior knowledge for robustness against outliers. • Bayesian risk is used to design the proposed method optimization problem. • The proposed method can be employed for big data using the AdaBoost approach. [ABSTRACT FROM AUTHOR]