Induction Algorithm Research Articles

An Ensemble Approach for Prioritizing Antivirals Against COVID-19 via Heterogeneous Network Inference-Based Inductive Matrix Completion

The global spread of COVID-19 caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) originated in Wuhan in December 2019, created a massive health crisis, and disrupted the world economy. Much research has been conducted to discover drugs, develop vaccines, and find repurposable drugs against the disease. Computational drug repurposing, the process of determining new uses for approved drugs through computational techniques, becomes an effective solution to fight the COVID-19 pandemic. This study aims to investigate and prioritize potential drugs against SARS-CoV-2 through an integrated network-based approach. We propose an ensemble approach based on network inference and inductive matrix completion (NIMCVDA) for virus–drug association prediction to identify antivirals against COVID-19. We constructed a heterogeneous drug–virus network using intra-similarities of virus genomic sequences and drug chemical structures and existing associations between viruses and drugs. A network inference method is used to infer missing drug–virus edges. Based on this, existing drug–virus association matrix is reconstructed. Finally, more accurate association scores between drugs and viruses are computed using the inductive matrix completion algorithm. The proposed method achieved an AUC of 0.9020 on five-fold cross-validation and 0.8786 on leave-one-out cross-validation. We compared the performance of the model with related approaches. In addition, we carried out case studies on the top-predicted drugs and implemented our model with other datasets to verify prediction performance. Our work prioritized repurposable drugs to battle with COVID-19 epidemic. The cross-validation results and case studies illustrate that the top-predicted drugs are strong candidates for further biological tests.

Sequential game theory based multi criterion network partitioning for controller placement in software defined wide area networks

The task of determining the number and location of controllers to be deployed in software defined network (SDN) affects various network performance metrics. Large-scale networks are typically divided into multiple domains of smaller size and then deploy one or more controller(s) in each domain. In this paper, we propose a sequential game theory based multi-criteria partitioning scheme for controller placement in large-scale SDN. The proposed scheme automatically identifies the optimal number of domains to be created in the network while considering multiple conflicting objectives. The proposed strategy comprises of three stages. The first stage generates the initial composition of domains. The second stage identifies the actual number of domains to be created in the network and the temporary composition of these domains by formulating several finite multi action two player sequential games. The third stage generates homogeneous domains by formulating several single action two player sequential games. The second and third stages are repeated until there is no improvement in the system objectives. We utilized a game tree for representing the finite sequential game. We also presented a backward induction algorithm for computing the player’s best response strategies, also known as Nash equilibrium. The proposed strategy is evaluated on medium and large-scale networks from Internet Topology Zoo. Further, the performance of the proposed approach was compared with three existing approaches from the literature. Although the complexity of the proposed SGTCP approach is asymptotically higher than existing approaches, the proposed approach converges rapidly in practice and the actual time it takes is quite low. The proposed approach outperforms existing state-of-the-art network partitioning approaches for controller placement in SDN.

Application of machine learning techniques for obesity prediction: a comparative study

Obesity, characterized by excess adipose tissue, is becoming a major public health problem. This condition, caused primarily by unbalanced energy intake (overconsumption) and exacerbated by modern lifestyles such as physical inactivity and suboptimal dietary habits, is the harbinger of a variety of health disorders such as diabetes, cardiovascular disease, and certain cancers. Therefore, there is an urgent need to accurately diagnose and assess the extent of obesity in order to formulate and apply appropriate preventive measures and therapeutic interventions. However, the heterogeneous results of existing diagnostic techniques have triggered a fierce debate on the optimal approach to identifying and assessing obesity, thus complicating the search for a standard diagnostic and treatment method. This research primarily aims to use machine learning techniques to build a robust predictive model for identifying overweight or obese individuals. The proposed model, derived from a person's physical characteristics and dietary habits, was evaluated using a number of machine learning algorithms, including Multilayer Perceptron (MLP), Support Vector Machine (SVM), Fuzzy K-Nearest Neighbors (FuzzyNN), Fuzzy Unordered Rule Induction Algorithm (FURIA), Rough Sets (RS), Random Tree (RT), Random Forest (RF), Naive Bayes (NB), Logistic Regression (LR), and Decision Table (DT). Subsequently, the developed models were evaluated using a number of evaluation measures such as correlation coefficient, accuracy, kappa statistic, mean absolute error, and mean square error. The hyperparameters of the model were properly calibrated to improve accuracy. The study revealed that the random forest model (RF) had the highest accuracy of 95.78 %, closely followed by the logistic regression model (LR) with 95.22 %. Other algorithms also produced satisfactory accuracy results but could not compete with the RF and LR models. This study suggests that the pragmatic application of the model could help physicians identify overweight or obese individuals and thus accelerate the early detection, prevention, and treatment of obesity-related diseases.

Separate-and-conquer survival action rule learning

Action mining is a data mining method that aims to identify recommendations for changing attribute values that can lead to the classification of data instances as examples of another class. Action mining algorithms extract rules containing recommendations in the premises and class changes in the conclusion. To the best of the authors’ knowledge, no method has been proposed yet for generating action rules based on censored data. This study introduces the first method for survival action rule generation. The method stems from the covering rule induction algorithm but generates rules defining the actions required to change not the class but the survival curve of the covered examples. Thus, this study poses a new research problem: generating action rules for censored data and survival analysis. This study evaluated the proposed method using 22 data sets in which two application domains of survival analysis were distinguished: medicine and predictive maintenance. In addition, more detailed analyses of the generated survival action rules were presented in the form of case studies for the two selected data sets. The results show that the proposed method generates good-quality survival action rules and changes in the survival curves, resulting from the identified actions are significant.

Software Requirement Risk Prediction Using Enhanced Fuzzy Induction Models

The development of most modern software systems is accompanied by a significant level of uncertainty, which can be attributed to the unanticipated activities that may occur throughout the software development process. As these modern software systems become more complex and drawn out, escalating software project failure rates have become a critical concern. These unforeseeable uncertainties are known as software risks, and they emerge from many risk factors inherent to the numerous activities comprising the software development lifecycle (SDLC). Consequently, these software risks have resulted in massive revenue losses for software organizations. Hence, it is imperative to address these software risks, to curb future software system failures. The subjective risk assessment (SRM) method is regarded as a viable solution to software risk problems. However, it is inherently reliant on humans and, therefore, in certain situations, imprecise, due to its dependence on an expert’s knowledge and experience. In addition, the SRM does not allow repeatability, as expertise is not easily exchanged across the different units working on a software project. Developing intelligent modelling methods that may offer more unbiased, reproducible, and explainable decision-making assistance in risk management is crucial. Hence, this research proposes enhanced fuzzy induction models for software requirement risk prediction. Specifically, the fuzzy unordered rule induction algorithm (FURIA), and its enhanced variants based on nested subset selection dichotomies, are developed for software requirement risk prediction. The suggested fuzzy induction models are based on the use of effective rule-stretching methods for the prediction process. Additionally, the proposed FURIA method is enhanced through the introduction of nested subset selection dichotomy concepts into its prediction process. The prediction performances of the proposed models are evaluated using a benchmark dataset, and are then compared with existing machine learning (ML)-based and rule-based software risk prediction models. From the experimental results, it was observed that the FURIA performed comparably, in most cases, to the rule-based and ML-based models. However, the FURIA nested dichotomy variants were superior in performance to the conventional FURIA method, and rule-based and ML-based methods, with the least accuracy, area under the curve (AUC), and Mathew’s correlation coefficient (MCC), with values of approximately 98%.

SHATTER: Searching Heterogeneous Combat Network Attack Sequences Through Network Embedding and Reinforcement Learning

A distributed combat system consists of various types of combat units interconnected through the network, which can be abstracted as a complex heterogeneous combat network (HCN). It is of great military value to study the disintegration of the operational capability of HCNs for operational mission planning and improving the survivability of operational networks. Heterogeneous networks can effectively fuse more information than homogeneous networks can. The function implementation of heterogeneous networks is also more complex. Finding the key nodes and obtaining an efficient attack sequence is our focus. Most current research on network disintegration has difficulty achieving efficiency and timeliness. In this article, we specifically propose searching heterogeneous combat network attack sequences through network embedding and reinforcement learning (SHATTER) model, which uses the induction algorithm combined with the graph neural network and reinforcement learning to solve the network disintegration sequence problem. Through experiments with synthetic heterogeneous network datasets and real heterogeneous networks, we determined that SHATTER significantly outperformed existing technologies.

Concise rule induction algorithm based on one-sided maximum decision tree approach

As the importance of machine learning tools for decision support continues to grow, interpretability has emerged as a key factor. Rule-based classification algorithms, such as decision trees and rule induction, enable high local interpretability by providing transparent reasoning rules in an IF-THEN format. In this context, it is essential to provide concise and clear rules and conditions to achieve high local interpretability. This study proposes a novel Concise Algorithm, designed to effectively remove irrelevant conditions from classification rules. We present a framework incorporating the Concise Algorithm, which employs the One-Sided-Maximum decision tree algorithm for rule generation, followed by the application of the Concise Algorithm to remove irrelevant conditions. This proposed framework produces a rule-based classification model that exhibits an enhanced predictive performance-interpretability trade-off compared to benchmark methods (CART, Ripper, CN2, and modified One-Sided-Maximum), as demonstrated by empirical tests conducted on 19 UCI datasets. A case study focusing on the breast-cancer-wisconsin dataset provides a comprehensive analysis of the rule and condition generation processes.

Big data decision tree for continuous-valued attributes based on unbalanced cut points

The decision tree is a widely used decision support model, which can quickly mine effective decision rules based on the dataset. The decision tree induction algorithm for continuous-valued attributes, based on unbalanced cut points, is efficient for mining decision rules; however, extending it to big data remains an unresolved. In this paper, two solutions are proposed to solve this problem: the first one is based on partitioning instance subsets, whereas the second one uses partitioning attribute subsets. The crucial of these two solutions is how to find the global optimal cut point from the set of local optimal cut points. For the first solution, the calculation of the Gini index of the cut points between computing nodes and the selection of the global optimal cut point by communication between these computing nodes is proposed. However, in the second solution, the division of the big data into subsets using attribute subsets in a way that all cut points of an attribute are on the same map node is proposed, the local optimal cut points can be found in this map node, then the global optimal cut point can be obtained by summarizing all local optimal cut points in the reduce node. Finally, the proposed solutions are implemented with two big data platforms, Hadoop and Spark, and compared with three related algorithms on four datasets. Experimental results show that the proposed algorithms can not only effectively solve the scalability problem, but also have lowest running time, the fastest speed and the highest efficiency under the premise of preserving the classification performance.

Fuzzy based inference system with ensemble classification based intrusion detection system in MANET

Mobile ad hoc networks (MANET) have become one of the hottest research areas in computer science, including in military and civilian applications. Such applications have formed a variety of security threats, particularly in unattended environments. An Intrusion detection system (IDS) must be in place to ensure the security and reliability of MANET services. These IDS must be compatible with the characteristics of MANETs and competent in discovering the biggest number of potential security threats. In this work, a specialized dataset for MANET is implemented to identify and classify three types of Denial of Service (DoS) attacks: Blackhole, Grayhole and Flooding Attack. This work utilized a cluster-based routing algorithm (CBRA) in MANET.A simulation to gather data, then processed to create eight attributes for creating a specialized dataset using Java. Mamdani fuzzy-based inference system (MFIS) is used to create dataset labelling. Furthermore, an ensemble classification technique is trained on the dataset to discover and classify three types of attacks. The proposed ensemble classification has six base classifiers, namely, C4.5, Fuzzy Unordered Rule Induction Algorithm (FURIA), Multilayer Perceptron (MLP), Multinomial Logistic Regression (MLR), Naive Bayes (NB) and Support Vector Machine (SVM). The experimental results demonstrate that MFIS with the Ensemble classification technique enables an enhancing security in MANET’s by modeling the interactions among a malicious node with number of legitimate nodes. This is suitable for future works on multilayer security problem in MANET.

Restricted Computations and Parameters in Type-Theory of Acyclic Recursion

The paper extends the formal language and the reduction calculus of Moschovakis type-theory of recursion, by adding a restrictor operator on terms with predicative restrictions. Terms with restrictions over memory variables formalise inductive algorithms with generalised, restricted parameters. The extended type-theory of restricted recursion (TTRR) provides computations for algorithmic semantics of mathematical expressions and definite descriptors, in formal and natural languages. The reduction calculi of TTRR provides a mathematical foundation of the work of compilers for reducing recursive programs to iterative ones. The type-theory of acyclic recursion (TTAR) has a special importance to syntax-semantics interfaces in computational grammars.

Discovering epistasis interactions in Alzheimer’s disease using integrated framework of ensemble learning and multifactor dimensionality reduction (MDR)

Alzheimer's disease (AD) is a complex disorder with strong genetic factors. The proposed framework is applied to Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. We present a novel framework integrating ensemble learning and MDR constructive induction algorithm to discover epistasis interactions associated with AD in a computationally efficient method. Discovering epistasis interactions is a big challenge and significantly impacts personalized medicine (PM). The applied ensemble learning algorithms are random forests (RF) with Gini index and permutation importance, Extreme Gradient Boosting (XGBoost), and classification and regression trees (CART). The classification accuracy of 5-way models varied between (0.8674–0.8758), whereas the accuracy of 2-way, 3-way, and 4-way models varied between (0.6515–0.6649), (0.7071–0.7170), and (0.7811–0.7878) respectively. The promising results of this proposed framework show high-ranked risk genes and up to 5-way epistasis models that contribute to the disease risk efficiently and at higher accuracy.

Introduction to the Special Section on FM 2021

The first attempts to apply the k-induction method to software verification are only recent. In this paper, we present a novel proof by induction algorithm, which is built on the top of a symbolic context-bounded model checker and uses an iterative ...

Evaluation of Emotions from Brain Signals on 3D VAD Space via Artificial Intelligence Techniques.

Recent achievements have made emotion studies a rising field contributing to many areas, such as health technologies, brain-computer interfaces, psychology, etc. Emotional states can be evaluated in valence, arousal, and dominance (VAD) domains. Most of the work uses only VA due to the easiness of differentiation; however, very few studies use VAD like this study. Similarly, segment comparisons of emotion analysis with handcrafted features also use VA space. At this point, we primarily focused on VAD space to evaluate emotions and segmentations. The DEAP dataset is used in this study. A comprehensive analytical approach is implemented with two sub-studies: first, segmentation (Segments I-VIII), and second, binary cross-comparisons and evaluations of eight emotional states, in addition to comparisons of selected segments (III, IV, and V), class separation levels (5, 4-6, and 3-7), and unbalanced and balanced data with SMOTE. In both sub-studies, Wavelet Transform is applied to electroencephalography signals to separate the brain waves into their bands (α, β, γ, and θ bands), twenty-four attributes are extracted, and Sequential Minimum Optimization, K-Nearest Neighbors, Fuzzy Unordered Rule Induction Algorithm, Random Forest, Optimized Forest, Bagging, Random Committee, and Random Subspace are used for classification. In our study, we have obtained high accuracy results, which can be seen in the figures in the second part. The best accuracy result in this study for unbalanced data is obtained for Low Arousal-Low Valence-High Dominance and High Arousal-High Valence-Low Dominance emotion comparisons (Segment III and 4.5-5.5 class separation), and an accuracy rate of 98.94% is obtained with the IBk classifier. Data-balanced results mostly seem to outperform unbalanced results.

Solving Finite-Horizon Discounted Non-Stationary MDPS

Abstract Research background Markov Decision Processes (MDPs) are a powerful framework for modeling many real-world problems with finite-horizons that maximize the reward given a sequence of actions. Although many problems such as investment and financial market problems where the value of a reward decreases exponentially with time, require the introduction of interest rates. Purpose This study investigates non-stationary finite-horizon MDPs with a discount factor to account for fluctuations in rewards over time. Research methodology To consider the fluctuations of rewards with time, the authors define new nonstationary finite-horizon MDPs with a discount factor. First, the existence of an optimal policy for the proposed finite-horizon discounted MDPs is proven. Next, a new Discounted Backward Induction (DBI) algorithm is presented to find it. To enhance the value of their proposal, a financial model is used as an example of a finite-horizon discounted MDP and an adaptive DBI algorithm is used to solve it. Results The proposed method calculates the optimal values of the investment to maximize its expected total return with consideration of the time value of money. Novelty No existing studies have before examined dynamic finite-horizon problems that account for temporal fluctuations in rewards.

Multi-dimensional Bayesian network classifiers for partial label ranking

The label ranking problem consists in learning preference models from training datasets labeled with (possibly incomplete) rankings of the class labels. The goal is then to predict a ranking for a given unlabeled instance. This work focuses on a more general interpretation where both the training dataset and the prediction given as output allow tied class labels, i.e., there is no particular preference between them. This problem is known as the partial label ranking problem. This paper tackles the partial label ranking problem by transforming the ranking with ties into a set of discrete variables representing the preference relations (ranked ahead, tied with, and ranked behind) between each pair of class labels. The posterior probabilities for each pair are then used to fill the values of a preference matrix. This preference matrix is the basis for solving the rank aggregation problem required to obtain the output ranking with ties. This paper aims to exploit the resemblance of this problem with multi-label and multi-dimensional classification by studying the use of Bayesian network classifiers to compute the posterior probabilities for the new class structure, i.e., pairs of class labels. In particular, binary relevance with naive Bayes and averaged one-dependence estimators between the new class structure are used to solve the partial label ranking problem. Furthermore, bivariate relationships between all the pairs of class labels are considered. However, the complexity of the model grows significantly, which makes it necessary to reduce the number of allowed bivariate relationships between pairs. Thus, a feature selection method is included to select the more relevant subset of bivariate relationships. The experimental evaluation shows that our proposals are competitive in accuracy with the current instance-based and decision tree induction algorithms. Moreover, they outperform the existing mixture-based probabilistic graphical models, while the algorithms proposed are much faster.

A Hybrid Model of Cancer Diseases Diagnosis Based on Gene Expression Data with Joint Use of Data Mining Methods and Machine Learning Techniques

One of the current focuses of modern bioinformatics is the development of hybrid models to process gene expression data, in order to create diagnostic systems for various diseases. In this study, we propose a solution to this problem that combines an inductive spectral clustering algorithm, random forest classifier, convolutional neural network, and alternative voting method for making the final decision about patient condition. In the first stage, we apply the spectral clustering algorithm to gene expression profiles using inductive methods of objective clustering, with the calculation of internal, external, and balance clustering quality criteria. This results in clusters of mutually correlated and differently expressed gene expression profiles. In the second stage, we apply the random forest classifier and convolutional neural network to identify the examined objects, containing as attributes the gene expression values in the allocated clusters. The presented research solves both binary- and multi-classification tasks. The final decision about the patient’s condition is made using the alternative voting method, considering the classification results based on the gene expression data in various clusters. The simulation results showed that the proposed technique was highly effective, achieving a high accuracy in object identification when both classifiers were used. However, the convolutional neural network had a significantly higher data processing efficiency than the random forest algorithm, due to its substantially shorter processing time.

'Krishivanam' An Innovation to Prevent the Damages Caused by Fire and Wild Animals to The Agriculture Land

Agricultural fields near forest areas often become the center of human-animal conflict. Due to increased deforestation and destruction of wild-life habitat, animals venture out of forest areas in search of food. Other major consequence of deforestation is forest fire and poaching. As solution to the issues specified, technology can help to detect and warn potential human-animal interference near forest areas. On detection of fire near the agricultural land, pump is turned on automatically to put off the fire. On detection of movement through IR sensor, camera is turned on to identify the animal or human intruder using Objective induction algorithm. If an animal is detected, animal specific sound repellent is played to drive the animal away. If any human with weapons is identified, then an alarm is raised to alert neighbors of possible poacher and intruder. The farmer is alerted by means of an alert message.

Exploring the interaction of cognition and emotion in small group collaborative discourse by Heuristic Mining Algorithm (HMA) and Inductive Miner Algorithm (IMA).

This study explored the interaction between cognition and emotion in blended collaborative learning. The participants (n = 30) of this study were undergraduate students enrolled in a 16-week course on information technology teaching. These students were divided into six groups of five people each. The behavior modes of the participants were analyzed using a heuristic mining algorithm and inductive miner algorithm. Compared with the groups with low task scores, the high-scoring groups exhibited more reflection phases and cycles in the interaction process and thus more frequent self-evaluation and regulation behavior for forethought and performance. Moreover, the frequency of emotion events unrelated with cognition was higher for the high-scoring groups than for the low-scoring groups. On the basis of the research results, this paper presents suggestions for developing online and offline blended courses.

PDF Malware Detection Based on Fuzzy Unordered Rule Induction Algorithm (FURIA)

The number of cyber-attacks is increasing daily, and attackers are coming up with new ways to harm their target by disseminating viruses and other malware. With new inventions and technologies appearing daily, there is a chance that a system might be attacked and its weaknesses taken advantage of. Malware is distributed through Portable Document Format (PDF) files, among other methods. These files’ adaptability makes them a prime target for attackers who can quickly insert malware into PDF files. This study proposes a model based on the Fuzzy Unordered Rule Induction Algorithm (FURIA) to detect PDF malware. The proposed model outperforms currently used methods in terms of reducing error rates and increasing accuracy. Other models, such as Naïve Bayes (NB), Decision Tree (J48), Hoeffding Tree (HT), and Quadratic Discriminant Analysis (QDA), were compared to the proposed model. The accuracy achieved by the proposed model is 99.81%, with an error rate of 0.0022.

Hepatitis C Virus prediction based on machine learning framework: a real-world case study in Egypt

Prediction and classification of diseases are essential in medical science, as it attempts to immune the spread of the disease and discover the infected regions from the early stages. Machine learning (ML) approaches are commonly used for predicting and classifying diseases that are precisely utilized as an efficient tool for doctors and specialists. This paper proposes a prediction framework based on ML approaches to predict Hepatitis C Virus among healthcare workers in Egypt. We utilized real-world data from the National Liver Institute, founded at Menoufiya University (Menoufiya, Egypt). The collected dataset consists of 859 patients with 12 different features. To ensure the robustness and reliability of the proposed framework, we performed two scenarios: the first without feature selection and the second after the features are selected based on sequential forward selection (SFS). Furthermore, the feature subset selected based on the generated features from SFS is evaluated. Naïve Bayes, random forest (RF), K-nearest neighbor, and logistic regression are utilized as induction algorithms and classifiers for model evaluation. Then, the effect of parameter tuning on learning techniques is measured. The experimental results indicated that the proposed framework achieved higher accuracies after SFS selection than without feature selection. Moreover, the RF classifier achieved 94.06% accuracy with a minimum learning elapsed time of 0.54 s. Finally, after adjusting the hyperparameter values of the RF classifier, the classification accuracy is improved to 94.88% using only four features.