Existing Learning Algorithms Research Articles

EMPD: An Efficient Membrane Potential Driven Supervised Learning Algorithm for Spiking Neurons

The brain-inspired spiking neurons, considered as the third generation of artificial neurons, are more biologically plausible and computationally powerful than traditional artificial neurons. One of the fundamental research in spiking neurons is to transform streams of incoming spikes into precisely timed spikes. Due to the inherent complexity of processing spike sequences, the formulation of efficient supervised learning algorithm is difficult and remains an important problem in the research area. This paper presents an efficient membrane potential driven (EMPD) supervised learning method capable of training neurons to generate desired sequences of spikes. The learning rule of EMPD is composed of two processes: 1) at desired output times, the gradient descent method is implemented to minimize the error function defined as the difference between the membrane potential and the firing threshold and 2) at undesired output time, synaptic weights are adjusted to make the membrane potential below the threshold. For efficiency, at undesired output times, EMPD calculates the membrane potential and makes a comparison with firing threshold only at some special time points when the neuron is most likely to cross the firing threshold. Experimental results show that the proposed EMPD approach has higher learning efficiency and accuracy over the existing learning algorithms.

Discriminative Representation Learning with Supervised Auto-encoder

Auto-encoders have been proved to be powerful unsupervised learning methods that able to extract useful features from input data or construct deep artificial neural networks by recent studies. In such settings, the extracted features or the initialized networks only learn the data structure while contain no class information which is a disadvantage in classification tasks. In this paper, we aim to leverage the class information of input to learn a reconstructive and discriminative auto-encoder. More specifically, we introduce a supervised auto-encoder that combines the reconstruction error and the classification error to form a unified objective function while taking the noisy concatenate data and label as input. The noisy concatenate input is constructed in such a method that one third has only original data and zero labels, one third has only label and zero data, the last one third has both original data and label. We show that the representations learned by the proposed supervised auto-encoder are more discriminative and more suitable for classification tasks. Experimental results demonstrate that our model outperforms many existing learning algorithms.

A novel method combining fuzzy SVM and sampling for imbalanced classification

The class imbalance problem has been reported to reduce performance of many existing learning algorithms in intrusion detection. However, the detection rates for minority classes still need to be improved. Thus, the novel hybrid method FSVMs is proposed to solve the problem in the paper, which integrates the prevailing sampling method SMOTE with fuzzy semi-supervised SVM learning approach to class imbalanced intrusion detection data. The basic KDD Cup 1999 dataset, NSLKDD dataset and imbalanced dataset from UCI are used to evaluate the performance of proposed model. Experiment results show that the proposed method outperforms other state-of-the-art classifiers including support vector machine (SVM), back propagation neural network (BPNN), Bayes, k-nearest neighbour (KNN), decision tree (DT), random forest (RF) and four sampling methods in the aspects of detection rate and false alarm rate, and has better robustness for imbalanced classification.

A Novel Method Combining Fuzzy SVM and Sampling for Imbalanced Classification

The class imbalance problem has been reported to reduce performance of many existing learning algorithms in intrusion detection. However, the detection rates for minority classes still need to be improved. Thus, the novel hybrid method FSVMs is proposed to solve the problem in the paper, which integrates the prevailing sampling method SMOTE with fuzzy semi-supervised SVM learning approach to class imbalanced intrusion detection data. The basic KDD Cup 1999 dataset, NSLKDD dataset and imbalanced dataset from UCI are used to evaluate the performance of proposed model. Experiment results show that the proposed method outperforms other state-of-the-art classifiers including support vector machine (SVM), back propagation neural network (BPNN), Bayes, k-nearest neighbour (KNN), decision tree (DT), random forest (RF) and four sampling methods in the aspects of detection rate and false alarm rate, and has better robustness for imbalanced classification.

A L-BFGS Based Learning Algorithm for Complex-Valued Feedforward Neural Networks

In this paper, a new learning algorithm is proposed for complex-valued feedforward neural networks (CVFNNs). The basic idea of this algorithm is that the descent directions of the cost function with respect to complex-valued parameters are calculated by limited-memory BFGS algorithm and the learning step is determined by Armijo line search method. Since the approximation of Hessian matrix is calculated by utilizing the information of the latest several iterations, the memory efficiency is improved. To keep away from the saturated ranges of activation functions, some gain parameters are adjusted together with weights and biases. Compared with some existing learning algorithms for CVFNNs, the convergence speed is faster and a deeper minima of the cost function can be reached by the developed algorithm. In addition, the effects of initial values of weights and biases on the efficiency and convergence speed of the learning algorithm are analyzed. The performance of the proposed algorithm is evaluated in comparison with some existing classifiers on a variety of benchmark classification problems. Experimental results show that better performance is achieved by our algorithm with relatively compact network structure.

Cloud robotics in Smart Manufacturing Environments: Challenges and countermeasures

In Smart Manufacturing Environments (SME), the use of cloud robotics is based on the integration of cloud computing and industrial robots, which provides a new technological approach to task execution and resource sharing compared to traditional industrial robots. However, research on cloud robotics in SME still faces some challenges. First, highly flexible load scheduling mechanisms are immature. Second, traditional optimization mechanisms for the network service quality do not meet the requirements of smart manufacturing due to time variability and service quality dynamics. And, finally, existing learning algorithms used without cloud-assisted resources cause great resource wasting. Accordingly, this paper explores main technologies related to cloud robotics in SME. The research contents include self-adaptive adjustment mechanisms for the service quality of a cloud robot network, computing load allocation mechanisms for cloud robotics, and group learning based on a cloud platform. The results presented in this paper are helpful to understand the internal mechanisms of perception and interaction, intelligent scheduling and control of cloud robot systems oriented to smart manufacturing, and the design of a cloud architecture oriented to group learning.

Neuroprosthetic Decoder Training as Imitation Learning.

Neuroprosthetic brain-computer interfaces function via an algorithm which decodes neural activity of the user into movements of an end effector, such as a cursor or robotic arm. In practice, the decoder is often learned by updating its parameters while the user performs a task. When the user’s intention is not directly observable, recent methods have demonstrated value in training the decoder against a surrogate for the user’s intended movement. Here we show that training a decoder in this way is a novel variant of an imitation learning problem, where an oracle or expert is employed for supervised training in lieu of direct observations, which are not available. Specifically, we describe how a generic imitation learning meta-algorithm, dataset aggregation (DAgger), can be adapted to train a generic brain-computer interface. By deriving existing learning algorithms for brain-computer interfaces in this framework, we provide a novel analysis of regret (an important metric of learning efficacy) for brain-computer interfaces. This analysis allows us to characterize the space of algorithmic variants and bounds on their regret rates. Existing approaches for decoder learning have been performed in the cursor control setting, but the available design principles for these decoders are such that it has been impossible to scale them to naturalistic settings. Leveraging our findings, we then offer an algorithm that combines imitation learning with optimal control, which should allow for training of arbitrary effectors for which optimal control can generate goal-oriented control. We demonstrate this novel and general BCI algorithm with simulated neuroprosthetic control of a 26 degree-of-freedom model of an arm, a sophisticated and realistic end effector.

Learning Behaviour Models of Discrete Event Production Systems from Observing Input/Output Signals

Learning behavior models out of event traces has been tackled in a wide variety of scientific projects and publications. Usually the resulting models are used for fault detection, reengineering, and analysis. But in practical applications, like monitoring, learned models can show high complexity and permissivity which makes it difficult to use these models and results tend to be ambiguous. Therefore, this paper defines so called Machine State Petri Nets (MSPN) with the aim of being generated out of recorded event traces and exploit additional information of the system to reduce the permissivity. An already existing learning algorithm has been extended by exploiting some facts common for most practical applications. An example shows how these adaptations improve the base algorithm regarding the aforementioned requirements.

An on-line weighted ensemble of regressor models to handle concept drifts

Many estimation, prediction, and learning applications have a dynamic nature. One of the most important challenges in machine learning is dealing with concept changes. Underlying changes may make the model designed on old data, inconsistent with new data. Also, algorithms usually specialize in one type of change. Other challenge is reusing previously acquired information in scenarios where changes may recur. This strategy improves the learning accuracy and reduces the processing time. Unfortunately, most existing learning algorithms to deal with changes are adapted on a batch basis. This process usually requires a long time, and such data may not reflect the current state of the system. However, even the system is adapted on a sample basis, existing algorithms may adapt slowly to changes and cannot conciliate old and new information. This paper proposes an On-line Weighted Ensemble (OWE) of regressor models which is able to learn incrementally sample by sample in the presence of several types of changes and simultaneously retain old information in recurring scenarios. The key idea is to keep a moving window that slides when a new sample is available. The error of each model on the current window is determined using a boosting strategy that assigns small errors to the models that predict accurately the samples predicted poorly by the ensemble. To handle recurring and non-recurring changes, OWE uses a new assignment of models׳ weights that takes into account the models׳ errors on the past and current windows using a discounting factor that decreases or increases the contribution of old windows. In addition, OWE launches new models if the system׳s accuracy is decreasing, and it can exclude inaccurate models over time. Experiments with artificial and industrial data reveal that in most cases OWE outperforms other state-of-the-art concept drift approaches.

Incremental Learning on a Budget and its Application to Quick Maximum Power Point Tracking of Photovoltaic Systems

Recent improvements in embedded systems has enabled learning algorithms to provide realistic solutions for system identification problems. Existing learning algorithms, however, continue to have limitations in learning on embedded systems, where physical memory space is constrained. To overcome this problem, we propose a Limited General Regression Neural Network (LGRNN), which is a variation of general regression neural network proposed by Specht or of simplified fuzzy inference systems. The LGRNN continues incremental learning even if the number of instances exceeds the maximum number of kernels in the LGRNN. We demonstrate LGRNN advantages by comparing it to other kernel-based perceptron learning methods. We also propose a light-weighted LGRNN algorithm, -LGRNNLight- for reducing computational complexity. As an example of its application, we present a Maximum Power Point Tracking (MPPT) microconverter for photovoltaic power generation systems. MPPT is essential for improving the efficiency of renewable energy systems. Although various techniques exist that can realize MPPT, few techniques are able to realize quick control using conventional circuit design. The LGRNN enables the MPPT converter to be constructed at low cost using the conventional combination of a chopper circuit and microcomputer control. The LGRNN learns the Maximum Power Point (MPP) found by Perturb and Observe (P&O), and immediately sets the converter reference voltage after a sudden irradiation change. By using this strategy, the MPPT quickly responds without a predetermination of parameters. The experimental results suggest that, after learning, the proposed converter controls a chopper circuit within 14 ms after a sudden irradiation change. This rapid response property is suitable for efficient power generation, even under shadow flicker conditions that often occur in solar panels located near large wind turbines.

Quality-Based Learning for Web Data Classification

The types of web data vary in terms of information quantity and quality. For example, some pages contain numerous texts, whereas some others contain few texts; some web videos are in high resolution, whereas some other web videos are in low resolution. As a consequence, the quality of extracted features from different web data may also vary greatly. Existing learning algorithms on web data classification usually ignore the variations of information quality or quantity. In this paper, the information quantity and quality of web data are described by quality-related factors such as text length and image quantity, and a new learning method is proposed to train classifiers based on quality-related factors. The method divides training data into subsets according to the clustering results of quality-related factors and then trains classifiers by using a multi-task learning strategy for each subset. Experimental results indicate that the quality-related factors are useful in web data classification, and the proposed method outperforms conventional algorithms that do not consider information quantity and quality.

Learning from Large Distributed Data: A Scaling Down Sampling Scheme for Efficient Data Processing

Extracting information from a training data set for predictive inference is a fundamental task in data mining and machine learning. With the exponential growth in the amount of data being generated in the past few years, there is an urgent need to develop or adapt existing learning algorithms to efficiently learn from large data sets. This paper describes three scaling techniques enabling machine learning algorithms to learn from large distributed data sets. First, a general single-pass formula for computing the covariance matrix of large data sets using the MapReduce framework is derived. Second, two new efficient and accurate sampling schemes for scaling down large data sets for local processing are presented. The first sampling scheme uses the single-pass covariance formula to select the most informative data points based on uncertainties in the linear discriminant score. The second technique on the other hand selects informative points based on uncertainties in the logistic regression model. A series of numerical experiments demonstrates numerically stable results from the application of the formula and a fast, efficient, accurate and cost effective sampling scheme.

Aircraft Engine Fuel Flow Prediction Using Process Neural Network

Monitoring the aircraft engine fuel flow is critical to the flight safety and the aircraft maintenance economy. Aim at predicting the aircraft engine fuel flow accurately and quickly, an aircraft engine fuel flow prediction method based on the process neural network is proposed in this paper. The learning speed of the existing learning algorithms (e.g. BP learning algorithm) for process neural network is too slow for the practical application. A Levenberg-Marquardt learning algorithm based on the expansion of the orthogonal basis functions is developed to raise the adaptability of the process neural network to the real problems. Finally, the proposed prediction method with the corresponding learning algorithm is utilized to predict the fuel flow of some aircraft engine, the results indicate that the proposed prediction method seems to perform well and appears suitable for using as an aircraft engine health condition monitoring tool, and the comparative results also indicate that the Levenberg-Marquardt learning algorithm has a faster learning convergence speed and a higher prediction accuracy than the BP learning algorithm.

Extreme Learning Machine-Based Classification of ADHD Using Brain Structural MRI Data

BackgroundEffective and accurate diagnosis of attention-deficit/hyperactivity disorder (ADHD) is currently of significant interest. ADHD has been associated with multiple cortical features from structural MRI data. However, most existing learning algorithms for ADHD identification contain obvious defects, such as time-consuming training, parameters selection, etc. The aims of this study were as follows: (1) Propose an ADHD classification model using the extreme learning machine (ELM) algorithm for automatic, efficient and objective clinical ADHD diagnosis. (2) Assess the computational efficiency and the effect of sample size on both ELM and support vector machine (SVM) methods and analyze which brain segments are involved in ADHD.MethodsHigh-resolution three-dimensional MR images were acquired from 55 ADHD subjects and 55 healthy controls. Multiple brain measures (cortical thickness, etc.) were calculated using a fully automated procedure in the FreeSurfer software package. In total, 340 cortical features were automatically extracted from 68 brain segments with 5 basic cortical features. F-score and SFS methods were adopted to select the optimal features for ADHD classification. Both ELM and SVM were evaluated for classification accuracy using leave-one-out cross-validation.ResultsWe achieved ADHD prediction accuracies of 90.18% for ELM using eleven combined features, 84.73% for SVM-Linear and 86.55% for SVM-RBF. Our results show that ELM has better computational efficiency and is more robust as sample size changes than is SVM for ADHD classification. The most pronounced differences between ADHD and healthy subjects were observed in the frontal lobe, temporal lobe, occipital lobe and insular.ConclusionOur ELM-based algorithm for ADHD diagnosis performs considerably better than the traditional SVM algorithm. This result suggests that ELM may be used for the clinical diagnosis of ADHD and the investigation of different brain diseases.

An Optimal PID Control Algorithm for Training Feedforward Neural Networks

The training problem of feedforward neural networks (FNNs) is formulated into a proportional integral and derivative (PID) control problem of a linear discrete dynamic system in terms of the estimation error. The robust control approach greatly facilitates the analysis and design of robust learning algorithms for multiple-input-multiple-output (MIMO) FNNs using robust control methods. The drawbacks of some existing learning algorithms can therefore be revealed clearly, and an optimal robust PID-learning algorithm is developed. The optimal learning parameters can be found by utilizing linear matrix inequality optimization techniques. Theoretical analysis and examples including function approximation, system identification, exclusive-or (XOR) and encoder problems are provided to illustrate the results.

IMPROVING MULTILABEL CLASSIFICATION BY AVOIDING IMPLICIT NEGATIVITY WITH INCOMPLETE DATA

Many real‐world problems require multilabel classification, in which each training instance is associated with a set of labels. There are many existing learning algorithms for multilabel classification; however, these algorithms assume implicit negativity, where missing labels in the training data are automatically assumed to be negative. Additionally, many of the existing algorithms do not handle incremental learning in which new labels could be encountered later in the learning process. A novel multilabel adaptation of the backpropagation algorithm is proposed that does not assume implicit negativity. In addition, this algorithm can, using a naïve Bayesian approach, infer missing labels in the training data. This algorithm can also be trained incrementally as it dynamically considers new labels. This solution is compared with existing multilabel algorithms using data sets from multiple domains, and the performance is measured with standard multilabel evaluation metrics. It is shown that our algorithm improves classification performance for all metrics by an overall average of 7.4% when at least 40% of the labels are missing from the training data and improves by 18.4% when at least 90% of the labels are missing.

A multi-stage learning framework for intelligent system

As information technologies advance and user-friendly interfaces develop, the interaction between humans and computers, information devices, and new consumer electronics is increasingly gaining attention. One example that most people can relate to is Apple's innovation in human-computer interaction which has been used on many products such as iPod and iPhone. Siri, the intelligent personal assistant, is a typical application of machine-learning human-computer interaction. Algorithms in machine learning have been employed in many disciplines, including gesture recognition, speaker recognition, and product recommendation systems. While the existing learning algorithms compute and learn from a large quantity of data, this study proposes an improved learning to rank algorithm named MultiStageBoost. In addition to ranking data through multiple stages, the MultiStageBoost algorithm significantly improves the existing algorithms in two ways. Firstly, it classifies and filters data to small quantities and applies the Boosting algorithm to achieve faster ranking performance. Secondly, it enhances the original binary classification by using the reciprocal of fuzzily weighted membership as the ranking distance. The importance of data is revealed in their ranked positions. Usually data ranked in the front are given more attention than those ranked in the middle. For example, after ranking 10,000 pieces of data, the top 10, or at most 100, are the most important and relevant. Whether the data after the top ones are ranked precisely does not really matter. Due to this reason, this study has made improvement on the conventional methods of the pair-wise ranking approach. Not only are data classified and ranked binarily, they are also given different weights depending on whether they are concordant or discordant. Incorporating the concept of weighting into the ranking distance allows us to increase the precision of ranking. Results from experiments demonstrate that our proposed algorithm outperforms the conventional methods in three evaluation measures: [email protected], MAP, and NDCG. MultiStageBoost was then applied to speech recognition. However, we do not aim to improve the technology of speech recognition, but simply hope to provide evidences that MultiStageBoost can be used in the classification and ranking in speech recognition. Experiments show that the recognition optimization procedures established by this study are able to increase the recognition rate to over 95% in the personal computing device and industrial personal computer. It is expected that in the future this voice management system will accurately and effectively identify speakers answering the voice response questionnaire and will successfully carry out the functions in the choice of answers, paying the way for the formation of a virtual customer service person.

Supervised Learning in Multilayer Spiking Neural Networks

We introduce a supervised learning algorithm for multilayer spiking neural networks. The algorithm overcomes a limitation of existing learning algorithms: it can be applied to neurons firing multiple spikes in artificial neural networks with hidden layers. It can also, in principle, be used with any linearizable neuron model and allows different coding schemes of spike train patterns. The algorithm is applied successfully to classic linearly nonseparable benchmarks such as the XOR problem and the Iris data set, as well as to more complex classification and mapping problems. The algorithm has been successfully tested in the presence of noise, requires smaller networks than reservoir computing, and results in faster convergence than existing algorithms for similar tasks such as SpikeProp.

Statistical models and learning algorithms for ordinal regression problems

In this study, we propose a learning algorithm for ordinal regression problems. In most existing learning algorithms, the threshold or location model is assumed to be the statistical model. For estimation of conditional probability of labels for a given covariate vector, we extended the location model to apply ordinal regressions. We present this learning algorithm using the squared-loss function with the location-scale models for estimating conditional probability. We prove that the estimated conditional probability satisfies the monotonicity of the distribution function. Furthermore, we have conducted numerical experiments to compare these proposed methods with existing approaches. We found that, in its ability to predict labels, our method may not have an advantage over existing approaches. However, for estimating conditional probabilities, it does outperform the learning algorithm using location models.

A modified learning algorithm for the multilayer neural network with multi-valued neurons based on the complex QR decomposition

In this paper, a modified learning algorithm for the multilayer neural network with the multi-valued neurons (MLMVN) is presented. The MLMVN, which is a member of complex-valued neural networks family, has already demonstrated a number of important advantages over other techniques. A modified learning algorithm for this network is based on the introduction of an acceleration step, performing by means of the complex QR decomposition and on the new approach to calculation of the output neurons errors: they are calculated as the differences between the corresponding desired outputs and actual values of the weighted sums. These modifications significantly improve the existing derivative-free backpropagation learning algorithm for the MLMVN in terms of learning speed. A modified learning algorithm requires two orders of magnitude lower number of training epochs and less time for its convergence when compared with the existing learning algorithm. Good performance is confirmed not only by the much quicker convergence of the learning algorithm, but also by the compatible or even higher classification/prediction accuracy, which is obtained by testing over some benchmarks (Mackey–Glass and Jenkins–Box time series) and over some satellite spectral data examined in a comparison test.