Feras Al-Obeidat

Differential Evolution for learning the classification method PROAFTN

This paper introduces a new learning technique for the multicriteria classification method PROAFTN. This new technique, called DEPRO, utilizes a Differential Evolution (DE) algorithm for learning and optimizing the output of the classification method PROAFTN. The limitation of the PROAFTN method is largely due to the set of parameters (e.g., intervals and weights) required to be obtained to perform the classification procedure. Therefore, a learning method is needed to induce and extract these parameters from data. DE is an efficient metaheuristic optimization algorithm based on a simple mathematical structure to mimic a complex process of evolution. Some of the advantages of DE over other global optimization methods are that it often converges faster and with more certainty than many other methods and it uses fewer control parameters. In this work, the DE algorithm is proposed to inductively obtain PROAFTNs parameters from data to achieve a high classification accuracy. Based on results generated from 12 public datasets, DEPRO provides excellent results, outperforming the most common classification algorithms.

An evolutionary framework using particle swarm optimization for classification method PROAFTN

Abstract: The aim of this paper is to introduce a methodology based on the particle swarm optimization (PSO) algorithm to train the Multi-Criteria Decision Aid (MCDA) method PROAFTN. PSO is an efficient evolutionary optimization algorithm using the social behavior of living organisms to explore the search space. It is a relatively new population-based metaheuristic that can be used to find approximate solutions to difficult optimization problems. Furthermore, it is easy to code and robust to control parameters. To apply PROAFTN, the values of several parameters need to be determined prior to classification, such as boundaries of intervals and weights. In this study, the proposed technique is named PSOPRO, which utilizes PSO to elicit the PROAFTN parameters from examples during the learning process. To test the effectiveness of the methodology and the quality of the obtained models, PSOPRO is evaluated on 12 public-domain datasets and compared with the previous work applied on PROAFTN. The computational results demonstrate that PSOPRO is very competitive with respect to the most common classification algorithms.

Automatic parameter settings for the PROAFTN classifier using hybrid particle swarm optimization

In this paper, a new hybrid metaheuristic learning algorithm is introduced to choose the best parameters for the classification method PROAFTN PROAFTN is a multi-criteria decision analysis (MCDA) method which requires values of several parameters to be determined prior to classification These parameters include boundaries of intervals and relative weights for each attribute The proposed learning algorithm, identified as PSOPRO-RVNS as it integrates particle swarm optimization (PSO) and Reduced Variable Neighborhood Search (RVNS), is used to automatically determine all PROAFTN parameters The combination of PSO with RVNS allows to improve the exploration capabilities of PSO by setting some search points to be iteratively re-explored using RVNS Based on the generated results, experimental evaluations show that PSOPRO-RVNS outperforms six well-known machine learning classifiers in a variety of problems.

A Fuzzy Decision Tree for Processing Satellite Images and Landsat Data

Abstract Satellite and airborne images, including Landsat, ASTER, and Hyperspectral data, are widely used in remote sensing and Geo- graphic Information Systems (GIS) to understand natural earth related processes, climate change, and anthropogenic activity. The nature of this type of data is usually multi or hyperspectral with individual spectral bands stored in raster file structures of large size and global coverage. The elevated number of bands (on the order of 200 to 250 bands) requires data processing algorithms capable of extracting information content, removing redundancy. Conventional statistical methods have been devised to reduce dimension- ality however they lack specific processing to handle data diversity. Hence, in this paper we propose a new data analytic technique to classify these complex multidimensional data cubes. Here, we use a well-known database consisting of multi-spectral values of pixels from satellite images, where the classification is associated with the central pixel in each neighborhood. The goal of our proposed approach is to predict this classification based on the given multi-spectral values. To solve this classification problem, we propose an improved decision tree (DT) algorithm based on a fuzzy approach. More particularly, we introduce a new hybrid classification algorithm that utilizes the conventional decision tree algorithm enhanced with the fuzzy approach. We propose an improved data classification algorithm that utilizes the best of a decision tree and multi-criteria classification. To investigate and evaluate the performance of our proposed method against other DT classifiers, a comparative and analytical study is conducted on well-known Landsat data.

Temperature Forecasts with Stable Accuracy in a Smart Home.

Abstract We forecast internal temperature in a home with sensors, modeled as a linear function of recent sensor values. When delivering forecasts as a service, two desirable properties are that forecasts have stable accuracy over a variety of forecast horizons – so service levels can be predicted – and that the forecasts rely on a modest amount of sensor history – so forecasting can be restarted soon after any data outage due to, for example, sensor failure. From a publicly available data set, we show that sensor values over the past one or two hours are sufficient to meet these demands. A standard machine learning method based on forward stepwise linear regression with cross validation gives forecasts whose out-of-sample errors increase slowly as the forecast horizon increases, and that are accurate to within one fifth of a degree C over three hours, and to within about one half degree C over six hours, based on one or two hours of history. Previous results from this data achieved errors within one degree C over three hours based on five days of history.

A multicriterion fuzzy classification method with greedy attribute selection for anomaly-based intrusion detection

Abstract Intrusion is widely recognized as a chronic and recurring problem of computer systems’ security with the continual changes and increasing volume of hacking techniques. This paper explores a new countermeasure approach for anomaly-based intrusion detection using a multicriterion fuzzy classification method combined with a greedy attribute selection. The proposed approach has the advantage of dealing with various types of attributes including network traffic basic TCP/IP packet headers, as well as content-based, time-based and host-based attributes. At the same time, to reduce the dimensionality and increase the computational efficiency, the greedy attribute selection algorithm enables it to choose an optimal subset of attributes that is most relevant for detecting intrusive events. The simplicity of the constructed model allows it to be replicated at various network components in emerging open system infrastructures such as sensor networks, wireless ad hoc networks, cloud computing, and smart grids. The proposed approach is evaluated and compared on a commonly-used intrusion detection benchmark dataset. The results show more than 99.9% overall accuracy with high detection rates for various types of intrusions can be achieved with about 26% only of the available attributes.

Opinions Sandbox: Turning Emotions on Topics into Actionable Analytics

The Opinions Sandbox is a running prototype that accesses comments collected from customers of a particular product or service, and calculates the overall sentiment toward that product or service. It performs topic extraction, displays the comments partitioned into topics, and presents a sentiment for each topic. This helps to quickly digest customers’ opinions, particularly negative ones, and sort them by the concerns expressed by the customers. These topics are now considered issues to be addressed. The Opinions Sandbox does two things with this list of issues. First, it simulates the social network of the future, after rectifying each issue. Comments with positive sentiment regarding this rectified issues are synthesized, they are injected into the comment corpus, and the effect on overall sentiment is produced. Second, it helps the user create a plan for addressing the issues identified in the comments. It uses the quantitative improvement of sentiment, calculated by the simulation in the first part, and it uses user-supplied cost estimates of the effort required to rectify each issue. Sets of possible actions are enumerated and analysed showing both the costs and the benefits. By balancing these benefits against these costs, it recommends actions that optimize the cost/benefit tradeoff.

Discretization Techniques and Genetic Algorithm for Learning the Classification Method PROAFTN

This paper introduces new techniques for learning the classification method PROAFTN from data. PROAFTN is a multi-criteria classification method and belongs to the class of supervised learning algorithms. To use PROAFTN for classification, some parameters must be obtained for this purpose. Therefore, an automatic method to extract these parameters from data with minimum classification errors is required. Here, discretization techniques and genetic algorithms are proposed for establishing these parameters and then building the classification model. Based on the obtained results, the newly proposed approach outperforms widely used classification methods.

Entity linking of tweets based on dominant entity candidates

Entity linking, also known as semantic annotation, of textual content has received increasing attention. Recent works in this area have focused on entity linking on text with special characteristics such as search queries and tweets. The semantic annotation of tweets is specially proven to be challenging given the informal nature of the writing and the short length of the text. In this paper, we propose a method to perform entity linking on tweets built based on one primary hypothesis. We hypothesize that while there are formally many possible entity candidates for an ambiguous mention in a tweet, as listed on the disambiguation page of the corresponding entity on Wikipedia, there are only few entity candidates that are likely to be employed in the context of Twitter. Based on this hypothesis, we propose a method to identify such dominant entity candidates for each ambiguous mention and use them in the annotation process. Particularly, our proposed work integrates two phases (i) dominant entity candidate detection, which applies community detection methods for finding the dominant candidates of ambiguous mentions; and (ii) named entity disambiguation that links a tweet to entities in Wikipedia by only considering the identified dominant entity candidates. Our investigations show that: (1) there are only very few entity candidates for each ambiguous mention in a tweet that need to be considered when performing disambiguation. This helps us limit the candidate search space and hence noticeably reduce the entity linking time; (2) limiting the search space to only a subset of disambiguation options will not only improve entity linking execution time but will also lead to improved accuracy of the entity linking process when the main entity candidates of each mention are mined from a temporally aligned corpus. We show that our proposed method offers competitive results with the state-of-the-art methods in terms of precision and recall on widely used gold standard datasets while significantly reducing the time for processing each tweet.

Neural word and entity embeddings for ad hoc retrieval

Abstract Learning low dimensional dense representations of the vocabularies of a corpus, known as neural embeddings, has gained much attention in the information retrieval community. While there have been several successful attempts at integrating embeddings within the ad hoc document retrieval task, yet, no systematic study has been reported that explores the various aspects of neural embeddings and how they impact retrieval performance. In this paper, we perform a methodical study on how neural embeddings influence the ad hoc document retrieval task. More specifically, we systematically explore the following research questions: (i) do methods solely based on neural embeddings perform competitively with state of the art retrieval methods with and without interpolation? (ii) are there any statistically significant difference between the performance of retrieval models when based on word embeddings compared to when knowledge graph entity embeddings are used? and (iii) is there significant difference between using locally trained neural embeddings compared to when globally trained neural embeddings are used? We examine these three research questions across both hard and all queries. Our study finds that word embeddings do not show competitive performance to any of the baselines. In contrast, entity embeddings show competitive performance to the baselines and when interpolated, outperform the best baselines for both hard and soft queries.