The value of Learning Rate and its change mechanisms is one of the issues in designing Learning systems such as Learning automata. In most cases a time-based reduction function is used to adjust the Learning Rate aim at reaching stability in training system. So the Learning Rate is a parameter that determines to what extent a Learning system is based on past experiences, and the impact of current events on it. This method is efficient but does not properly function in dynamic and non-stationary environments.In this paper, a new method for Adaptive Learning Rate adjustment in Learning automata is proposed. In this method, in addition to the length of time to learn, some statistical characteristics of actions probability vector of Learning Automata are used to determine the increase or decrease of Learning Rate. Furthermore, unlike existing methods, during the process of Learning, both increase and decrease of the Learning Rate is done and Learning Automata responds effectively to changes in the dynamic random environment.Empirical studies show that the proposed method has more flexibility in compatibility to the non-stationary dynamic environments and get out of local maximum points and the learned values are closer to the true values.

This article investigates the problem of simultaneous attitude and vibration control of a flexible spacecraft to perform high precision attitude maneuvers and reduce vibrations caused by the flexible panel excitations in the presence of external disturbances, system uncertainties, and actuator faults. Adaptive integral sliding mode control is used in conjunction with an attitude actuator fault iterative Learning observer (based on sliding mode) to develop an active fault tolerant algorithm considering rigid-flexible body dynamic interactions. The discontinuous structure of fault-tolerant control led to discontinuous commands in the control signal, resulting in chattering. This issue was resolved by introducing an Adaptive rule for the sliding surface. Furthermore, the utilization of the sign function in the iterative Learning observer for estimating actuator faults has not only enhanced its robustness to external disturbances through a straightforward design, but has also led to a decrease in computing workload. The strain Rate feedback control algorithm has been employed with the use of piezoelectric sensor/actuator patches to minimize residual vibrations caused by rigid-flexible body dynamic interactions and the effect of attitude actuator faults. Lyapunov's law ensures finite-time overall system stability even with fully coupled rigid-flexible nonlinear dynamics. Numerical simulations demonstRate the performance and advantages of the proposed system compared to other conventional approaches.

Distance-based clustering methods categorize samples by optimizing a global criterion, finding ellipsoid clusters with roughly equal sizes. In contrast, density-based clustering techniques form clusters with arbitrary shapes and sizes by optimizing a local criterion. Most of these methods have several hyper-parameters, and their performance is highly dependent on the hyper-parameter setup. Recently, a Gaussian Density Distance (GDD) approach was proposed to optimize local criteria in terms of distance and density properties of samples. GDD can find clusters with different shapes and sizes without any free parameters. However, it may fail to discover the appropriate clusters due to the interfering of clustered samples in estimating the density and distance properties of remaining unclustered samples. Here, we introduce Adaptive GDD (AGDD), which eliminates the inappropriate effect of clustered samples by Adaptively updating the parameters during clustering. It is stable and can identify clusters with various shapes, sizes, and densities without adding extra parameters. The distance metrics calculating the dissimilarity between samples can affect the clustering performance. The effect of different distance measurements is also analyzed on the method. The experimental results conducted on several well-known datasets show the effectiveness of the proposed AGDD method compared to the other well-known clustering methods.

The convergence of Learning Rate in neural networks identifier and controller is one of challenging issues which attracts great interest from researchers. This paper suggests the Adaptive gradient descent algorithm with Learning laws which assures the convergence of multi-layer perceptron neural network based on Taylor series expansion of output error. In the proposed method the Learning Rate can be calculated online. To increase the accuracy and the speed of convergence, the second and higher order terms of the Taylor series expansion are not considered constant and are updated during the algorithm. Simulating the suggested algorithm on two examples reveals that with considering the bounds in the proposed method, the aims for Learning Rate, convergence of Learning algorithm are guaranteed and the speed of convergence of training algorithm is increased.

Fraud refers to earn wealth including property, goods and services through immoral and non-legal channels. Fraud has always been in action and experiences an increasing trend worldwide. Fraud in financial transactions not only leads to losing huge financial resources, but also leads to reduction in trust of customers on using modern banking systems and hence, reduction in efficiency of the systems and optimal management of financial transactions. In recent years, by emerging new technologies of banking industry, new means of fraud are discovered. Although a new information system carry advantages and benefits, new opportunities are made for fraudsters. The applications of fraud detection methods encompasses detection of frauds in an organization, analysis of frauds and also user/customer behavior analytics in order to predict future behavior and reduce the fraud risks. In recent decades, employing new technologies in management of banking transactions has risen. Banks and financial institutions inevitably migRated from traditional banking to modern online banking to provide effective services. Although, the use of online banking systems improves the management of financial processes and speeds up services to customers of institutions, but some issues would also be carried. Financial frauds is one of the issues which organizations seek to prevent and reduce effects. In this paper, a novel machine Learning based model is presented to detect fraud in electronic banking transactions using profile data of bank customers. In the proposed method, transactional data from banks are leveraged and a multi-layer perceptron neural network with Adaptive Learning Rate is trained to prove the validity of a transaction and hence, improve the fraud detection in electronic banking. The proposed method shows promising results compared with logistic regression and support vector machines.