This study proposes a novel machine learning architecture and various learning algorithms to build-in anti-phishing services for avoiding cyber-phishing attack. For the rapid develop of information technology, hackers engage in cyber-phishing attack to steal important personal information, which draws information security concerns. The prevention of phishing website involves in various aspect, for example, user training, public awareness, fraudulent phishing, etc. However, recent phishing research has mainly focused on preventing fraudulent phishing and relied on manual identification that is inefficient for real-time detection systems. In this study, we used methods such as ANOVA, X2, and information gain to evaluate features. Then, we filtered out the unrelated features and obtained the top 28 most related features as the features to use for the training and evaluation of traditional machine learning algorithms, such as Support Vector Machine (SVM) with linear or rbf kernels, Logistic Regression (LR), Decision tree, and K-Nearest Neighbor (KNN). This research also evaluated the above algorithms with the ensemble learning concept by combining multiple classifiers, such as Adaboost, bagging, and voting. Finally, the eXtreme Gradient Boosting (XGBoost) model exhibited the best performance of 99.2%, among the algorithms considered in this study.

In this letter, we propose two robust and distributed game-based algorithms, which are the modifications of two algorithms proposed in [1], to solve the joint base station selection and resource allocation problem with imperfect information in heterogeneous cellular networks (HCNs). In particular, we repeatedly sample the received payoffs in the exploitation stage of each algorithm to guarantee the convergence when the payoffs of some users (UEs) in [1] cannot accurately be acquired for some reasons. Then, we derive the rational sampling number and prove the convergence of the modified algorithms. Finally, simulation results demonstrate that two modified algorithms achieve good convergence performances and robustness in the incomplete information scheme.

The minimized assumption generation has been recognized as an important improvement of the assume-guarantee verification method in order to generate minimal assumptions. The generated minimal assumptions can be used to recheck the whole component-based software at a lower computational cost. The method is not only fitted to component-based software but also has a potential to solve the state space explosion problem in model checking. However, the computational cost for generating the minimal assumption is very high so the method is difficult to be applied in practice. This paper presents an optimization as a continuous work of the minimized assumption generation method in order to reduce the complexity of the method. The key idea of this method is to find a smaller assumption in a sub-tree of the search tree containing the candidate assumptions using the depth-limited search strategy. With this approach, the improved method can generate assumptions with a lower computational cost and consumption memory than the minimized method. The generated assumptions are also effective for rechecking the systems at much lower computational cost in the context of software evolution. An implemented tool supporting the improved method and experimental results are also presented and discussed.

This letter investigates an iterative learning control with advanced output data (ADILC) scheme for non-minimum phase (NMP) systems when the number of NMP zeros is unknown. ADILC has a simple learning structure that can be applied to both minimum phase and NMP systems. However, in the latter case, it is assumed that the number of NMP zeros is already known. In this paper, we propose an ADILC scheme in which the number of NMP zeros is unknown. Based on input-to-output mapping, the learning starts from the relative degree. When the input becomes larger than a certain upper bound, we redesign the input update law which consists of the relative degree and the estimated value for the number of NMP zeros.

It is important to optimize aggregation schemes for heterogeneous wireless networks for maximizing communication throughput utilizing any available radio access networks. In the heterogeneous networks, differences of the quality of service (QoS), such as throughput, delay and packet loss rate, of the networks makes difficult to maximize the aggregation throughput. In this paper, we firstly analyze influences of such differences in QoS to the aggregation throughput, and show that it is possible to improve the throughput by adjusting the parameters of an aggregation system. Since manual parameter optimization is difficult and takes much time, we propose an autonomous parameter tuning scheme using a machine learning algorithm for the heterogeneous wireless network aggregation. We implement the proposed scheme on a heterogeneous cognitive radio network system. The results on our experimental network with network emulators show that the proposed scheme can improve the aggregation throughput better than the conventional schemes. We also evaluate the performance using public wireless network services, such as HSDPA, WiMAX and W-CDMA, and verify that the proposed scheme can improve the aggregation throughput by iterating the learning cycle even for the public wireless networks. Our experimental results show that the proposed scheme achieves twice better aggregation throughput than the conventional schemes.

An assume-guarantee verification method has been recognized as a promising approach to verify component-based software by model checking. This method is not only fitted to component-based software but also has a potential to solve the state space explosion problem in model checking. The method allows us to decompose a verification target into components so that we can model check each of them separately. In this method, assumptions are seen as the environments needed for the components to satisfy a property and for the rest of the system to be satisfied. The number of states of the assumptions should be minimized because the computational cost of model checking is influenced by that number. Thus, we propose a method for generating minimal assumptions for the assume-guarantee verification of component-based software. The key idea of this method is finding the minimal assumptions in the search spaces of the candidate assumptions. The minimal assumptions generated by the proposed method can be used to recheck the whole system at much lower computational cost. We have implemented a tool for generating the minimal assumptions. Experimental results are also presented and discussed.

FeedForward (FF-) Blind Source Separation (BSS) systems have some degree of freedom in the solution space. Therefore, signal distortion is likely to occur. First, a criterion for the signal distortion is discussed. Properties of conventional methods proposed to suppress the signal distortion are analyzed. Next, a general condition for complete separation and distortion-free is derived for multi-channel FF-BSS systems. This condition is incorporated in learning algorithms as a distortion-free constraint. Computer simulations using speech signals and stationary colored signals are performed for the conventional methods and for the new learning algorithms employing the proposed distortion-free constraint. The proposed method can well suppress signal distortion, while maintaining a high source separation performance.

Over the years, many improvements and refinements to the backpropagation learning algorithm have been reported. In this paper, a new adaptive penalty-based learning extension for the backpropagation learning algorithm and its variants is proposed. The new method initially puts pressure on artificial neural networks in order to get all outputs for all training patterns into the correct half of the output range, instead of mainly focusing on minimizing the difference between the target and actual output values. The upper bound of the penalty values is also controlled. The technique is easy to implement and computationally inexpensive. In this study, the new approach is applied to the backpropagation learning algorithm as well as the RPROP learning algorithm. The superiority of the new proposed method is demonstrated though many simulations. By applying the extension, the percentage of successful runs can be greatly increased and the average number of epochs to convergence can be well reduced on various problem instances. The behavior of the penalty values during training is also analyzed and their active role within the learning process is confirmed.

We consider the classification problem to construct a classifier c:{0,1}n{0,1} from a given set of examples (training set), which (approximately) realizes the hidden oracle y:{0,1}n{0,1} describing the phenomenon under consideration. For this problem, a number of approaches are already known in computational learning theory; e.g., decision trees, support vector machines (SVM), and iteratively composed features (ICF). The last one, ICF, was proposed in our previous work (Haraguchi et al., (2004)). A feature, composed of a nonempty subset S of other features (including the original data attributes), is a Boolean function fS:{0,1}S{0,1} and is constructed according to the proposed rule. The ICF algorithm iterates generation and selection processes of features, and finally adopts one of the generated features as the classifier, where the generation process may be considered as embodying the idea of boosting, since new features are generated from the available features. In this paper, we generalize a feature to an extended Boolean function fS:{0,1,*}S{0,1,*} to allow partial knowledge, where * denotes the state of uncertainty. We then propose the algorithm ICF* to generate such generalized features. The selection process of ICF* is also different from that of ICF, in that features are selected so as to cover the entire training set. Our computational experiments indicate that ICF* is better than ICF in terms of both classification performance and computation time. Also, it is competitive with other representative learning algorithms such as decision trees and SVM.

This paper describes a new fast learning algorithm for Self-Organizing Map employing a "rough comparison winner-take-all" and its digital hardware architecture. In rough comparison winner-take-all algorithm, the winner unit is roughly and strictly assigned in early and later learning stage, respectively. It realizes both of high accuracy and fast learning. The digital hardware of the self-organizing map with proposed WTA algorithm is implemented using FPGA. Experimental results show that the designed hardware is superior to other hardware with respect to calculation speed.

We present a learning algorithm of the Hopfield neural network for minimizing edge crossings in linear drawings of nonplanar graphs. The proposed algorithm uses the Hopfield neural network to get a local optimal number of edge crossings, and adjusts the balance between terms of the energy function to make the network escape from the local optimal number of edge crossings. The proposed algorithm is tested on a variety of graphs including some "real word" instances of interconnection networks. The proposed learning algorithm is compared with some existing algorithms. The experimental results indicate that the proposed algorithm yields optimal or near-optimal solutions and outperforms the compared algorithms.

We propose an efficient algorithm for making multi-layered neural networks (MLN) fault-tolerant to all multiple weight faults in a multi-dimensional interval by injecting intentionally two extreme multi-dimensional values in the interval into the weights of the selected multiple links in a learning phase. The degree of fault-tolerance to a multiple weight fault is measured by the number of essential multiple links. First, we analytically discuss how to choose effectively the multiple links to be injected, and present a learning algorithm for making MLNs fault tolerant to all multiple (i.e., simultaneous) faults in the interval defined by two multi-dimensional extreme points. Then it is proved that after the learning algorithm successfully finishes, MLNs become fault tolerant to all multiple faults in the interval. It is also shown that the time in a weight modification cycle depends little on multiplicity of faults k for small k. These are confirmed by simulation.

This paper presents an efficient Hybrid Learning Algorithm (HLA) for Radial Basis Function Neural Network (RBFNN). The HLA combines the gradient method and the linear least squared method for adjusting the RBF parameters and connection weights. The number of hidden neurons and their characteristics are determined using an unsupervised clustering procedure, and are used as input parameters to the learning algorithm. We demonstrate that the HLA, while providing faster convergence in training phase, is also less sensitive to training and testing patterns. The proposed HLA in conjunction with RBFNN is used as a classifier in a face recognition system to show the usefulness of the learning algorithm. The inputs to the RBFNN are the feature vectors obtained by combining shape information and Pseudo Zernike Moment (PZM). Simulation results on the Olivetti Research Laboratory (ORL) database and comparison with other algorithms indicate that the HLA yields excellent recognition rate with less hidden neurons in human face recognition.

This paper attempts to establish a theory for a general auto-associative memory model. We start by defining a new concept called supporting function to replace the concept of energy function. As known, the energy function relies on the assumption of symmetric interconnection weights, which is used in the conventional Hopfield auto-associative memory, but not evidenced in any biological memories. We then formulate the information retrieving process as a dynamic system by making use of the supporting function and derive the attraction or asymptotic stability condition and the condition for convergence of an arbitrary state to a desired state. The latter represents a key condition for associative memory to have a capability of learning from variant samples. Finally, we develop an algorithm to learn the asymptotic stability condition and an algorithm to train the system to recover desired states from their variant samples. The latter called sample learning algorithm is the first of its kind ever been discovered for associative memories. Both recalling and learning processes are of finite convergence, a must-have feature for associative memories by analogy to normal human memory. The effectiveness of the recalling and learning algorithms is experimentally demonstrated.

In this paper, three issues concerning the linear adaptive receiver using the LMS algorithm for single-user demodulation in direct-sequence/code-division multiple-access (DS/CDMA) systems are considered. First, the convergence rate of the LMS algorithm in DS/CDMA environment is considered theoretically. Both upper and lower bounds of the eigenvalue spread of the autocorrelation matrix of receiver input signals are derived. It is cleared from the results that the convergence rate of the LMS algorithm becomes slow when the signal power of interferer is large. Second, fast converging technique using a prefilter is considered. The LMS based adaptive receiver using an adaptive prefilter adjusted by a Hebbian learning algorithm to decorrelate the input signals is proposed. Computer simulation results show that the proposed receiver provides faster convergence than the LMS based receiver. Third, the complexity reduction of the proposed receiver by prefiltering is considered. As for the reduced complexity receiver, it is shown that the performance degradation is little as compared with the full complexity receiver.

A noise reduction filter composed of a sandglass-type neural network (Sandglass-type Neural network Noise Reduction Filter: SNNRF) was proposed in the present paper. Sandglass-type neural network (SNN) has symmetrical layer construction, and consists of the same number of units in input and output layers and less number of units in a hidden layer. It is known that SNN has the property of processing signals which is equivalent to KL expansion after learning. We applied the recursive least square (RLS) method to learning of SNNRF, so that the SNNRF became able to process on-line noise reduction. This paper showed theoretically that SNNRF behaves most optimally when the number of units in the hidden layer is equal to the rank of covariance matrix of signal component included in input signal. Computer experiments confirmed that SNNRF acquired appropriate characteristics for noise reduction from input signals, and remarkably improved the SN ratio of the signals.

A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.

Recently, many researches on associative memories have been made a lot of neural network models have been proposed. Bidirectional Associative Memory (BAM) is one of them. The BAM uses Hebbian learning. However, unless the traning vectors are orthogonal, Hebbian learning does not guarantee the recall of all training pairs. Namely, the BAM which is trained by Hebbian learning suffers from low memory capacity. To improve the storage capacity of the BAM, Pseudo-Relaxation Learning Algorithm for BAM (PRLAB) has been proposed. However, PRLAB needs long learning epochs because of random initial weights. In this paper, we propose Quick Learning for BAM which greatly reduces learning epochs and guarantees the recall of all training pairs. In the proposed algorithm, the BAM is trained by Hebbian learning in the first stage and then trained by PRLAB. Owing to the use of Hebbian learning in the first stage, the weights are much closer to the solution space than the initial weights chosen randomly. As a result, the proposed algorithm can reduce the learning epocks. The features of the proposed algorithm are: 1) It requires much less learning epochs. 2) It guarantees the recall of all training pairs. 3) It is robust for noisy inputs. 4) The memory capacity is much larger than conventional BAM. In addition, we made clear several important chracteristics of the conventional and the proposed algorithms such as noise reduction characteristics, storage capacity and the finding of an index which relates to the noise reduction.

A new regularization cost function for generalization in real-valued function learning is proposed. This cost function is derived from the maximum likelihood method using a modified sample distribution, and consists of a sum of square errors and a stabilizer which is a function of integrated square derivatives. Each of the regularization parameters which gives the minimum estimation error can be obtained uniquely and non-empirically. The parameters are not constants and change in value during learning. Numerical simulation shows that this cost function predicts the true error accurately and is effective in neural network learning.

Analysis of satellite images requires classificatio of image objects. Since different categories may have almost the same brightness or feature in high dimensional remote sensing data, many object categories overlap with each other. How to segment the object categories accurately is still an open question. It is widely recognized that the assumptions required by many classification methods (maximum likelihood estimation, etc.) are suspect for textural features based on image pixel brightness. We propose an image feature based neural network approach for the segmentation of AVHRR images. The learning algoriothm is a modified backpropagation with gain and weight decay, since feedforward networks using the backpropagation algorithm have been generally successful and enjoy wide popularity. Destructive algorithms that adapt the neural architecture during the training have been developed. The classification accuracy of 100% is reached for a validation data set. Classification result is compared with that of Kohonen's LVQ and basic backpropagation algorithm based pixel-by-pixel method. Visual investigation of the result images shows that our method can not only distinguish the categories with similar signatures very well, but also is robustic to noise.