Number Of Possible Features Research Articles

Exploring Feature Selection Scenarios for Deep Learning-based Side-channel Analysis

One of the main promoted advantages of deep learning in profiling sidechannel analysis is the possibility of skipping the feature engineering process. Despite that, most recent publications consider feature selection as the attacked interval from the side-channel measurements is pre-selected. This is similar to the worst-case security assumptions in security evaluations when the random secret shares (e.g., mask shares) are known during the profiling phase: an evaluator can identify points ofinterest locations and efficiently trim the trace interval. To broadly understand how feature selection impacts the performance of deep learning-based profiling attacks, this paper investigates three different feature selection scenarios that could be realistically used in practical security evaluations. The scenarios range from the minimum possible number of features (worst-case security assumptions) to the whole available traces. Our results emphasize that deep neural networks as profiling models show successful key recovery independently of explored feature selection scenarios against first-order masked software implementations of AES-128. First, we show that feature selection with the worst-case security assumptions results in optimal profiling models that are highly dependent on the number of features and signal-to-noise ratio levels. Second, we demonstrate that attacking raw side-channel measurements with small deep neural networks also provides optimal models, that shortens the gap between worst-case security evaluations and online (realistic) profiling attacks. In all explored feature selection scenarios, the hyperparameter search always indicates a successful model with up to eight hidden layers for MLPs and CNNs, suggesting that complex models are not required for the considered datasets. Our results demonstrate the key recovery with less than ten attack traces for all datasets for at least one of the feature selection scenarios. Additionally, in several cases, we can recover the target key with a single attack trace.

Selection of optimal wavelet features for epileptic EEG signal classification with LSTM

Epilepsy remains one of the most common chronic neurological disorders; hence, there is a need to further investigate various models for automatic detection of seizure activity. An effective detection model can be achieved by minimizing the complexity of the model in terms of trainable parameters while still maintaining high accuracy. One way to achieve this is to select the minimum possible number of features. In this paper, we propose a long short-term memory (LSTM) network for the classification of epileptic EEG signals. Discrete wavelet transform (DWT) is employed to remove noise and extract 20 eigenvalue features. The optimal features were then identified using correlation and P value analysis. The proposed method significantly reduces the number of trainable LSTM parameters required to attain high accuracy. Finally, our model outperforms other proposed frameworks, including popular classifiers such as logistic regression (LR), support vector machine (SVM), K-nearest neighbor (K-NN) and decision tree (DT).

A hybrid mine blast algorithm for feature selection problems

Feature selection (FS) is the process of finding the least possible number of features that are able to describe a dataset in the same way as the original features. Feature selection is a crucial preprocessing step for data mining techniques as it improves the performance of the prediction process in terms of speed and accuracy and also provides a better understanding of stored data. The success of the FS process depends on achieving a balance between two important factors, namely selecting the minimal number of features and maintaining the maximum accuracy in the results. In this paper, two methods are proposed to improve the FS process. Firstly, the mine blast algorithm (MBA) is introduced to optimize the FS process in the exploration phase. Secondly, the MBA is hybridized with simulated annealing as a local search in the exploitation phase to enhance the solutions located by the MBA. The proposed approaches (MBA and MBA–SA) are tested on 18 benchmark datasets from the UCI repository, and the comprehensive experimental results indicate that MBA–SA achieved good performance when compared with five approaches in the literature.

Collaboration and followership: A stochastic model for activities in social networks.

In this work we investigate how future actions are influenced by the previous ones, in the specific contexts of scientific collaborations and friendships on social networks. We describe the activity of the agents, providing a model for the formation of the bipartite network of actions and their features. Therefore we only require to know the chronological order in which the actions are performed, and not the order in which the agents are observed. Moreover, the total number of possible features is not specified a priori but is allowed to increase along time, and new actions can independently show some new-entry features or exhibit some of the old ones. The choice of the old features is driven by a degree-fitness method: indeed, the probability that a new action shows one of the old features does not solely depend on the popularity of that feature (i.e. the number of previous actions showing it), but it is also affected by some individual traits of the agents or the features themselves, synthesized in certain quantities, called fitnesses or weights, that can have different forms and different meaning according to the specific setting considered. We show some theoretical properties of the model and provide statistical tools for the parameters’ estimation. The model has been tested on three different datasets and the numerical results are provided and discussed.

Vulnerability Analysis of Instructions for SDC-Causing Error Detection

Due to the centralization of communication in the management of data generated by diverse Internet of Thing (IoT) devices, there is a lack of reliability when data is being transferred and stored. Among errors caused by various behaviors, Silent Data Corruption (SDC) error, owing to stealthy destruction without error prompt, is one of the most difficult data consistency problems in the storage system, whether it is a traditional multi-control, distributed storage, or public cloud storage. Nowadays, for SDC error detection, extracting instruction features to analyze vulnerabilities of programs or instructions has still not been fully explored. Literature generally just count the number of possible features, without mining the inter-characteristic of the instruction and correlation between them. Thus, we propose a method of SDC-causing Error Detection based on Support Vector Regression (SED-SVR) for fully exploiting the correlation between data features. Specifically, firstly, we extract instruction features based on the SDC vulnerability of program instructions by analyzing results of fault injections. Secondly, we establish the instruction SDC vulnerability prediction model based on SVR and propose our SED-SVR model. Thirdly, according to the predicted values of SDC vulnerability, we develop some solutions for faults tolerance of target programs by different granularity of instruction redundancy. The experimental results show that our SED-SVR has higher fault detection rate with lower performance overhead.

Classification and Prediction of Brain Disorders Using Functional Connectivity: Promising but Challenging.

Brain functional imaging data, especially functional magnetic resonance imaging (fMRI) data, have been employed to reflect functional integration of the brain. Alteration in brain functional connectivity (FC) is expected to provide potential biomarkers for classifying or predicting brain disorders. In this paper, we present a comprehensive review in order to provide guidance about the available brain FC measures and typical classification strategies. We survey the state-of-the-art FC analysis methods including widely used static functional connectivity (SFC) and more recently proposed dynamic functional connectivity (DFC). Temporal correlations among regions of interest (ROIs), data-driven spatial network and functional network connectivity (FNC) are often computed to reflect SFC from different angles. SFC can be extended to DFC using a sliding-window framework, and intrinsic connectivity states along the time-varying connectivity patterns are typically extracted using clustering or decomposition approaches. We also briefly summarize window-less DFC approaches. Subsequently, we highlight various strategies for feature selection including the filter, wrapper and embedded methods. In terms of model building, we include traditional classifiers as well as more recently applied deep learning methods. Moreover, we review representative applications with remarkable classification accuracy for psychosis and mood disorders, neurodevelopmental disorder, and neurological disorders using fMRI data. Schizophrenia, bipolar disorder, autism spectrum disorder (ASD), attention deficit hyperactivity disorder (ADHD), Alzheimer's disease and mild cognitive impairment (MCI) are discussed. Finally, challenges in the field are pointed out with respect to the inaccurate diagnosis labeling, the abundant number of possible features and the difficulty in validation. Some suggestions for future work are also provided.

Sequential and Mixed Genetic Algorithm and Learning Automata (SGALA, MGALA) for Feature Selection in QSAR.

Feature selection is of great importance in Quantitative Structure-Activity Relationship (QSAR) analysis. This problem has been solved using some meta-heuristic algorithms such as GA, PSO, ACO and so on. In this work two novel hybrid meta-heuristic algorithms i.e. Sequential GA and LA (SGALA) and Mixed GA and LA (MGALA), which are based on Genetic algorithm and learning automata for QSAR feature selection are proposed. SGALA algorithm uses advantages of Genetic algorithm and Learning Automata sequentially and the MGALA algorithm uses advantages of Genetic Algorithm and Learning Automata simultaneously. We applied our proposed algorithms to select the minimum possible number of features from three different datasets and also we observed that the MGALA and SGALA algorithms had the best outcome independently and in average compared to other feature selection algorithms. Through comparison of our proposed algorithms, we deduced that the rate of convergence to optimal result in MGALA and SGALA algorithms were better than the rate of GA, ACO, PSO and LA algorithms. In the end, the results of GA, ACO, PSO, LA, SGALA, and MGALA algorithms were applied as the input of LS-SVR model and the results from LS-SVR models showed that the LS-SVR model had more predictive ability with the input from SGALA and MGALA algorithms than the input from all other mentioned algorithms. Therefore, the results have corroborated that not only is the predictive efficiency of proposed algorithms better, but their rate of convergence is also superior to the all other mentioned algorithms.

Intrusion detection systems using a hybrid SVD-based feature extraction method

Intrusion detection systems (IDSs) are able to diagnose network anomalies with the help of machine learning techniques. This paper presents a novel singular value decomposition (SVD)-based method that creates a new feature, which is applied to an IDS. The main goal is to build an effective model on datasets, which have the least possible number of features. Using the least possible number of features is inevitable in case of improving the efficiency and de-escalating the effect of curse of dimensionality in datasets with large number of features. The proposed method combines the SVD method with four classification algorithms; decision tree, Naive Bayes, neural networks and SVM, to obtain a high accuracy in anomaly detection. This method is applied on the KDD CUP 99 and NSL_KDD datasets. Results of simulations indicate that the proposed method provides a considerable improvement in accuracy, compared with ordinary feature selection methods.

Modeling Networks with a Growing Feature-Structure

We present a new network model accounting for multidimensional assortativity. Each node is characterized by a number of features and the probability of a link between two nodes depends on common features. We do not fix a priori the total number of possible features. The bipartite network of the nodes and the features evolves according to a stochastic dynamics that depends on three parameters that respectively regulate the preferential attachment in the transmission of the features to the nodes, the number of new features per node, and the power-law behavior of the total number of observed features. Our model also takes into account a mechanism of triadic closure. We provide theoretical results and statistical estimators for the parameters of the model. We validate our approach by means of simulations and an empirical analysis of a network of scientific collaborations.

A network model characterized by a latent attribute structure with competition

The quest for a model that is able to explain, describe, analyze and simulate real-world complex networks is of uttermost practical, as well as theoretical, interest. In fact, networks can be a natural way to represent many phenomena; often, they arise from a complex interweaving of some features of the nodes. For example, in a co-authorship network, a link stems more easily between authors with similar interests; similarly, in a genetic regulatory network, links are affected by the different biological functions of the regulators.In this paper we introduce and study a novel network model that is based on a latent attribute structure: this model, inspired by a generalization of the Indian Buffet process, is simple and contains a small number of parameters, with a clear and intuitive role. Each node is characterized by a number of features and the probability of the existence of an edge between two nodes depends on the features they share; the number of possible features is not fixed a priori and can grow indefinitely. Moreover, a random fitness parameter is introduced for each node in order to determine its ability to transmit its own features to other nodes; this behavior is added on top of a process of Indian-Buffet type. Because of the fitness property, a node’s connectivity does not depend on its age alone, so that also “young but fit” nodes are able to compete and succeed in propagating their features and acquiring links. We also show how, considering the resulting bipartite node-attribute network, it is possible to gain some insight about which nodes were originally the most “fit”.Our model for this bipartite network depends on few parameters, that are characterized by their straightforward interpretation and by the availability of proper estimators. Even if the parameters are easy to interpret and tune, the model is general enough to represent complex phenomena—e.g., homophily, heterophily, or any interplay between features. We provide some theoretical as well as experimental results regarding the power-law behavior of the model and the proposed tools for the estimation of the parameters. We also show, through a number of experiments, how the proposed model naturally captures most local and global properties (e.g., degree distributions, connectivity and distance distributions) real networks exhibit.

Estimating the Difficulty of Touchless Hand Gestures

As part of a Natural User Interface we can use human body gestures, although they must be evaluated to get better results. We can evaluate them using a quantitative model. For this purpose, we begin the analysis of touchless hand gestures based on significant gesture features. From the large number of possible features we selected the length and the directions of trajectory. We propose to use the quantity of distance units and directions for describing and quantifying gestures. Moreover, we found that well-defined changes of directions (or corners) influence the speed for performing a gesture, thus users need more time to execute the gesture. Additionally, a model was developed using these three features. The model allows estimating the difficulty of gestures before users testing.

Comparative study of decision tree classifier and best first tree classifier for fault diagnosis of automobile hydraulic brake system using statistical features

Hydraulic brakes are the most important components in automobile. It requires advanced supervision and fault diagnosis to improve the safety of passenger, reliability and economy. Condition monitoring is one of the major division through which the reliability of such components could be monitored. The condition of the brake components can be monitored by using the vibration characteristics which will reveal the condition of the brake systems. In this paper machine learning algorithm using vibration monitoring is proposed as a possible solution to this problem. From the hydraulic brake test set up, the vibration signals were acquired by using a piezoelectric transducer and data acquisition system. C4.5 decision tree algorithm was used to extract statistical features from vibration signals. Feature selection was also carried out. Since no much of methodologies are available to find the effective number of features for a given problem, a detailed study is needed to find the best possible number of features. Hence the effect of number of features was studied by using decision tree. The selected features were classified using C4.5 decision tree algorithm and Best first decision tree algorithm with pre pruning and post pruning techniques. The results are discussed and conclusions of the study are presented.

On user-defined features

Feature-based design is becoming one of the fundamental design paradigms of CAD systems. In this paradigm, the basic unit is a feature and parts are constructed by a sequence of feature attachment operations. The type and number of possible features involved depend upon product type, the application reasoning process and the level of abstraction. Therefore to provide CAD systems with a basic mechanism to define features that fit the end-user needs seems more appropriate than trying to provide a large repertoire of features covering every possible application.A procedural mechanism is proposed for generating and deploying user-defined features in a feature-based design paradigm. The usefulness of the mechanism relies on two functional capabilities. First the shape and size of the user-defined features are instantiated according to parameter values given by the end-user. Second the end-user positions and orients the feature in the part being designed by means of geometric gestures on geometric references.