UCI Repository Datasets Research Articles

Improve the effectiveness of machine learning models in detecting website phishing using morphological features in URL analysis

With the proliferation of the Internet, the emergence of various threats has become increasingly prevalent, particularly the danger posed by phishing websites. These websites are designed with malicious content aimed at exploiting users who inadvertently access them. This method of attack represents a significant potential risk for users in cyberspace. The problem of detecting and eliminating phishing websites has garnered significant interest and research within the community. In this study, we propose a set of morphological features in URL path analysis, combined with machine learning methods, to detect phishing website URLs. Experimental evaluation with the UCI Repository dataset results have demonstrated the effectiveness of the proposed feature set in terms of all metrics (Accuracy, Precision, Recall, and F1 Score) compared to previous methods.

Heart Disease Prediction Using Machine Learning

The heart plays an important part in a living creature. The opinion and prognosis of a heart complaint must be made easily, exhaustively, and directly, because the slightest negligence can lead to serious complications or death. Numerous heart conditions are risk factors for death, and the number is gradually increasing. To solve this problem, prophetic styles that will ameliorate people's understanding of the complaint are urgently demanded. Machine literacy is a part of AI known for providing predictive support for any situation that requires training from natural wonders. In this, we compute the fineness of ML algorithms for cardiac prognostication, similar to k-nearest neighbor, decision tree, direct retrogression, and support vector machines, through training and evaluation using the UCI repository dataset (SVM). Anaconda (Jupytor) Primer is the stylish tool to use Python programming. It has colorful functions in the library and title lines to make it more effective and accessible.

A stable variant of linex loss SVM for handling noise with reduced hyperparameters

Support Vector Machine (SVM) primarily uses the hinge loss function with maximum margin. The boundary instances determine the separating hyperplanes. However, in real-world scenarios, noisy instances around the decision boundary can degrade the performance of hinge SVM. To address this issue, researchers introduced the pinball SVM, which uses all the instances to determine the separating hyperplane and has a discontinuous piecewise linear loss function. A more recent variant, the linex loss SVM with NAG, employs a combination of linear and exponential loss functions, resulting in an asymmetric and continuous loss function. Due to the employment of Nesterov accelerated gradient (NAG), the older linex SVM formulation was unstable. This study presents a novel noise-robust variant of linex SVM, termed linexSVM, which is stable in performance compared to the linex SVM with NAG. Additionally, the proposed linexSVM has fewer hyperparameters in the dual formulation than other noise-robust SVM variants, such as pinball SVM, ϵ-insensitive pinball SVM, and ϵ1,ϵ2 pinball SVM. We evaluated the performance of the proposed linexSVM, hinge SVM, and pinball SVM using the UCI dataset repository. The experimental results indicate that the proposed linexSVM outperforms hinge and pinball loss SVMs on datasets with and without noise.

Sistem Klasifikasi Penyakit Jantung Menggunakan Teknik Pendekatan SMOTE Pada Algoritma Modified K-Nearest Neighbor

The heart is a vital organ that plays a crucial role in pumping oxygenated blood and nutrients throughout the body. Heart disease refers to damage to the heart that can occur in various forms, caused by infections or congenital abnormalities. The World Health Organization (WHO) reports nearly 17.9 million deaths each year due to heart disease. In Indonesia, the prevalence of heart disease is around 1.5%, meaning that in 2018, approximately 15 out of 1,000 people, or nearly 2,784,060 individuals, were affected by this disease, according to the Basic Health Research data (Riskesdas) 2018. Many people have limited knowledge about heart health, leading to a lack of awareness of their heart conditions. This can be attributed to a lack of understanding regarding the importance of medical checkups related to heart health. Modified K-Nearest Neighbors (MKNN) is one of the data mining methods applied for classifying the risk of heart disease. The research utilized data obtained from the UCI dataset repository, which consists of 918 records with 12 attributes. To balance the imbalanced dataset with minority classes, the Synthetic Minority Over-sampling Technique (SMOTE) approach was used to generate new synthetic samples from the minority class. The objective of developing a web-based system for heart disease classification is to assist the public in assessing their risk of heart disease as early as possible, enabling them to take preventive actions sooner. The accuracy results of the MKNN algorithm with a 90:10 ratio are 80.37%, while with the MKNN+SMOTE approach, the accuracy increased to 84.00%. The use of the SMOTE approach improved the accuracy of low-performing data.

A DISEASE PREDICTION MODEL USING SPOTTED HYENA SEARCH OPTIMIZATION AND BI-LSTM

Early disease prediction is the best solution for any deadly disease in this fast internet world. In this direction, machine learning and deep learning techniques are applied in this fast world to predict diseases early and update the disease level through the Internet of Things techniques. For this purpose, this paper proposes a disease prediction system that uses the newly proposed fuzzy temporal correlation aware classifier and the auto-encoded bidirectional long short-term memory (FTC-Bi-LSTM) to predict the diseases such as heart, cancer, and diabetes. Moreover, this paper proposes a hybrid optimization algorithm called spotted hyena and cuckoo search optimization algorithm (SHCSA) to select the contributed features used to enrich the prediction accuracy. The standard benchmark UCI repository dataset is used to conduct the various experiments and obtain better performance in terms of precision, recall, f-measure, and prediction accuracy. The result is validated by considering the hospital patient data as input and proving the performance.

A DISEASE PREDICTION MODEL USING SPOTTED HYENA SEARCH OPTIMIZATION AND BI-LSTM

Early disease prediction is the best solution for any deadly disease in this fast internet world. In this direction, machine learning and deep learning techniques are applied in this fast world to predict diseases early and update the disease level through the Internet of Things techniques. For this purpose, this paper proposes a disease prediction system that uses the newly proposed fuzzy temporal correlation aware classifier and the auto-encoded bidirectional long short-term memory (FTC-Bi-LSTM) to predict the diseases such as heart, cancer, and diabetes. Moreover, this paper proposes a hybrid optimization algorithm called spotted hyena and cuckoo search optimization algorithm (SHCSA) to select the contributed features used to enrich the prediction accuracy. The standard benchmark UCI repository dataset is used to conduct the various experiments and obtain better performance in terms of precision, recall, f-measure, and prediction accuracy. The result is validated by considering the hospital patient data as input and proving the performance.

A Quasi-Oppositional Based Flamingo Search Algorithm Integrated with Generalized Ring Crossover for Effective Feature Selection

ABSTRACT The goal of this article is to present a new hybrid technique for solving the feature selection problem. Conventionally, the process of determining the most relevant subset based on the given criteria is known as feature selection. Specifically, feature selection is a real world issue that can be tackled by an optimization technique. In the proposed method, a novel hybrid Quasi oppositional based Flamingo search algorithm with Generalized Ring Crossover (QOFSA-GRC) model is introduced to pick the most relevant features from the dataset. Specifically, the Quasi oppositional based Flamingo search algorithm (QOFSA) generates two populations one by Quasi oppositional based learning and the other by the Flamingo search algorithm to resolve the issue known as the curse of dimensionality. Then, by utilizing generalized ring crossover, the multiple relevant features are selected from the UCI repository dataset. Finally, the Kernel Extreme Learning Machine (KELM) classifier validates the selected features. The performance of the proposed model is tested by 20 UCI benchmark datasets and the outcomes are compared with other models. Through experimental outcomes, it has been revealed that the suggested model produces the best accuracy and also selects fewer numbers of features that are more relevant from the dataset. In terms of other performance measures, the proposed model attains better outcomes for the majority of the datasets.

Coupling of Dimensionality Reduction and Stacking Ensemble Learning for Smartphone-Based Human Activity Recognition

Human activity recognition (HAR) plays a vital role in the field of ambient assisted living (AAL) for the welfare of the elders who live alone in the home. AAL provides service through ambient sensors, vision systems, smartphone devices, and wearable sensors. Smartphone devices are familiar, portable, cost-effective, and make the process of monitoring easier. Various research works have proposed smartphone-based HAR systems to recognize basic and complex activities. However, the results are not satisfactory for the case of postural transitions such as stand-to-sit, sit-to-sleep, etc. To improve the recognition rate, this paper couples principal component analysis with stacking ensemble learning for dimensionality reduction and classification respectively. Extensive experimentation of UCI repository datasets such as UCI-HAR has been performed and the performances are measured using familiar metrics such as accuracy, precision, and recall.

Small-sample continual learning classification method with vaccine to update memory cells based on the artificial immune system

In this paper, a novel continual learning classification method (SCLM) in small sample cases is proposed, which inspired by the immune system's continuous improvement of immunity through injecting vaccines. Data-driven classification method requires a large number of historical data to establish a pattern recognition model with good generalization performance. However, in practice, the data that can be used for training is usually small and unbalanced, which lead to poor classification accuracy. In addition, batch learning method cannot improve continually classification performance by learning test phase data. In view of the above problems, SCLM generates sample as vaccine by finding the group center of training samples, so that B cells mature and activate memory cells in the train phase. In the test phase, the recognition ability of SCLM is further improved by learning new samples and updating memory cells. In order to evaluate its performance under the condition of less training samples and its possible advantages, the experiments on well-known datasets in UCI repository and reciprocating compressor faults diagnose were performed. The results show that SCLM has better classification performance than other methods when the number of training samples is insufficient. At the same time, the method of generating data has significantly improved the classification performance of other methods.

Rainfall Prediction using Machine Learning

Abstract: In India, Agriculture contributes major role to Indian economy. For agriculture, Rainfall is important but during these days’ rainfall prediction has become a major challenging problem. Good prediction of rainfall provides knowledge and know in advance to take precautions and have better strategy about theirs crops. Global warming is also having severe effect on nature as well as mankind and it accelerates the change in climatic conditions. Because of its air is getting warmer and level of ocean is rising, leads to flood and cultivated field is changing into drought. Due to adverse climatic change leads to unseasonable and unreasonable amount of rainfall. To predict Rainfall is one of the best techniques to know about rainfall and climate. The main aim of this study revolves around providing correct climate description to the clients from various perspectives like agriculture, researchers, generation of power etc. to grasp the need of transformation in climate and its parameters like temperature, humidity, precipitation, wind speed that eventually directs to projection of rainfall. Rainfall also depends on geographic locations hence is an arduous task to predict. Machine Learning is the evolving subset of an AI, that helps in predicting the rainfall. In this research paper, we will be using UCI repository dataset with multiple attributes for predicting the rainfall. The main aim of this study is to develop the rainfall prediction system and predict the rainfall with better accuracy with the use of Machine Learning classification algorithms. Keywords: Rainfall Prediction system, Machine Learning, Dataset, Classification algorithms.

Hybrid slime mould - Grey wolf optimization algorithm for efficient feature selection

Selection of features is an effective method for minimizing the amount of data features in order to improve machine learning classification performance. Choosing a set of attributes is a high-level procedure for selecting a collection of relevant features. To boost the classifier's performance, use a dimensional dataset. We outline a typical feature selection problem in this work in order to minimize the amount of role and responsibilities while improving accuracy. Different classification dataset from the Machine learning repository have been used to test SMA with GWO as a feature selection strategy. The feature selections for UCI repository datasets include Bat optimization, Cuckoo search optimization, Slime mould optimization, Whale optimization, Particle swarm optimization, and Grey wolf optimization.

ORBoost: An Orthogonal AdaBoost

Ensemble learners and deep neural networks are state-of-the-art schemes for classification applications. However, deep networks suffer from complex structure, need large amount of samples and also require plenty of time to be converged. In contrast, ensemble learners (especially AdaBoost) are fast to be trained, can work with small and large datasets and also benefit strong mathematical background. In this paper, we have developed a new orthogonal version of AdaBoost, termed as ORBoost, in order to desensitize its performance against noisy samples as well as exploiting low number of weak learners. In ORBoost, after reweighting the distribution of each learner, the Gram-Schmidt rule updates those weights to make a new samples’ distribution to be orthogonal to the former distributions. In contrast in AdaBoost, there is no orthogonality constraint even between two successive weak learners while there is a similarity between the distributions of samples in different learners. To assess the performance of ORBoost, 16 UCI-Repository datasets along with six big datasets are deployed. The performance of ORBoost is compared to the standard AdaBoost, LogitBoost and AveBoost-II over the selected datasets. The achieved results support the significant superiority of ORBoost to the counterparts in terms of accuracy, robustness, number of exploited weak learners and generalization on most of the datasets.

An evolutionary algorithm based feature selection and fuzzy rule reduction technique for the prediction of skin cancer

SummaryIn current years, the death rate from skin cancers (SCs) tends to develop pretty. Various research verified that SC rank third as a deadliest disease, after breast and lung cancer. It will become vital to diagnose this malignancy at an early stage. The objective of this research is to mix machine learning and soft computing techniques to gain higher accuracy within the prediction of SC. To play out the exploration work, we utilized two data sets, one from “Save Life Hospital,” India, and the other is the UCI repository skin cancer data set. In this article, three meta‐heuristic algorithms, the FS_GA, the FS_PSO, and the FS_ACO, were used to select the best features from the data set provided to it. The AFRG_algorithm generates a set of fuzzy rules automatically and the RR_algorithm reduces certain fuzzy rules from the fuzzy system. For the SCC_dataset, the end accuracy obtained was 97.67%, 98.45%, and 99.22%. For the UCI_dataset, the end accuracy obtained was 98.81%, 99.72%, and 99.67%. Experimental results on the used datasets show that the proposed method strikingly improves the forecast exactitude of skin malignancy.

Benchmarking Analysis of the Accuracy of Classification Methods Related to Entropy.

In the machine learning literature we can find numerous methods to solve classification problems. We propose two new performance measures to analyze such methods. These measures are defined by using the concept of proportional reduction of classification error with respect to three benchmark classifiers, the random and two intuitive classifiers which are based on how a non-expert person could realize classification simply by applying a frequentist approach. We show that these three simple methods are closely related to different aspects of the entropy of the dataset. Therefore, these measures account somewhat for entropy in the dataset when evaluating the performance of classifiers. This allows us to measure the improvement in the classification results compared to simple methods, and at the same time how entropy affects classification capacity. To illustrate how these new performance measures can be used to analyze classifiers taking into account the entropy of the dataset, we carry out an intensive experiment in which we use the well-known J48 algorithm, and a UCI repository dataset on which we have previously selected a subset of the most relevant attributes. Then we carry out an extensive experiment in which we consider four heuristic classifiers, and 11 datasets.

Improved Cuckoo Search Algorithm for Future Selection

The classification of high-dimensional dataset is challenging as it contains large amount irrelevant and noisy features. Thus, feature selection is performed in the dataset to eliminate these redundant features. It reduces the dimensionality of the dataset and increases the classification accuracy. Hence, for selecting the relevant features in high dimensional data, an improved cuckoo search algorithm (ICSA) was proposed in this paper. After feature selection, the dataset undergo classification using KNN classifier and SVM classifier. The experimental process illustrates that the improved cuckoo search algorithm effectively increases the classification accuracy by reducing the number of features in the dataset. For analysing the proposed algorithm, seven UCI repository dataset have been utilised. Also, the ICS algorithm is compared with other existing algorithms for the given dataset. From the investigation process, it was concluded that the proposed algorithm selects lesser number of features and also enhances the classification accuracy than the other existing algorithms.

Hybrid Heart Disease Prediction Supervised, Unsupervised Opinion Mining Algorithms

Heart plays significant role in living organisms. Diagnosis and prediction of heart related diseases requires more precision, perfection and correctness because slightly mistake can cause fatigue problem or death of the person, there are numerous death cases related to heart and their counting is increasing exponentially day by day. To affect the matter there's essential need of prediction system for awareness about diseases Machine learning is that the branch of AI (AI), it provides prestigious support in predicting any quite event which take training from natural events. During this paper, we calculate accuracy of machine learning algorithms for predicting heart condition, for this algorithms are k-nearest neighbor, decision tree, linear regression and support vector machine (SVM) by using UCI repository dataset for training and testing. For implementation of Python programming Anaconda (jupytor) notebook is best tool, which have many kind of library, header file, that make the work more accurate and precise.

A Hybrid Whale Optimization Algorithm for Global Optimization

This paper proposes a hybrid whale optimization algorithm (WOA) that is derived from the genetic and thermal exchange optimization-based whale optimization algorithm (GWOA-TEO) to enhance global optimization capability. First, the high-quality initial population is generated to improve the performance of GWOA-TEO. Then, thermal exchange optimization (TEO) is applied to improve exploitation performance. Next, a memory is considered that can store historical best-so-far solutions, achieving higher performance without adding additional computational costs. Finally, a crossover operator based on the memory and a position update mechanism of the leading solution based on the memory are proposed to improve the exploration performance. The GWOA-TEO algorithm is then compared with five state-of-the-art optimization algorithms on CEC 2017 benchmark test functions and 8 UCI repository datasets. The statistical results of the CEC 2017 benchmark test functions show that the GWOA-TEO algorithm has good accuracy for global optimization. The classification results of 8 UCI repository datasets also show that the GWOA-TEO algorithm has competitive results with regard to comparison algorithms in recognition rate. Thus, the proposed algorithm is proven to execute excellent performance in solving optimization problems.

Weighted support vector machine using fuzzy rough set theory

The existence of both uncertainty and imprecision has detrimental impact on efficiency of decision-making applications and some machine learning methods, in particular support vector machine in which noisy samples diminish the performance of SVM training. Therefore, it is important to introduce a special method in order to improve this problem. Fuzzy aspects can handle mentioned problem which has been considered in some classification methods. This paper presents a novel weighted support vector machine to improve the noisy sensitivity problem of standard support vector machine for multiclass data classification. The basic idea is considered to add a weighted coefficient to the penalty term Lagrangian formula for optimization problem, which is called entropy degree, using lower and upper approximation for membership function in fuzzy rough set theory. As a result, noisy samples have low degree and important samples have high degree. To evaluate the power of the proposed method WSVM-FRS (Weighted SVM-Fuzzy Rough Set), several experiments have been conducted based on tenfold cross-validation over real-world data sets from UCI repository and MNIST data set. Experimental results show that the proposed method is superior than the other state-of-the-art competing methods regarding accuracy, precision and recall metrics.

Breast Cancer Diagnosis Using Cluster-based Undersampling and Boosted C5.0 Algorithm

Learning from imbalanced data set is relatively new challenge for breast cancer diagnosis, where the diseases cases are often quite rare relative to normal population. Although traditional algorithms are all accuracy-oriented which result biased towards the majority class. The combinations of sampling methods with ensemble classifiers have shown certainly good performance. In this paper, a hybrid of cluster-based undersampling and boosted C5.0 is proposed. The proposed classification model consists of two phases: cluster analysis and classification. In cluster analysis, affinity propagation algorithm is used to define the number of clusters, and then the k-means clustering is utilized to select the border and informative samples. In the classification phase, C5.0 algorithm is used in conjunction with boosting technical, owing to leverage the strength of the individual classifiers. The proposed algorithm is assessed by 14 benchmark imbalanced data sets taken from UCI dataset repository. The extensive experimental results on different imbalanced datasets demonstrated that the proposed algorithm can achieve better classification performance in terms of Matthews’ Correlation Coefficient (MCC) as compared to other existing imbalanced dataset classification algorithms.

The linguistic summarization and the interpretability, scalability of fuzzy representations of multilevel semantic structures of word-domains

ABSTRACT The effect of the linguistic (L-) summarization mined from a given dataset D by a human-made method M strongly depends on the fuzzy sets constructed to represent the L-words of dataset attributes. One can observe that the semantics of words is objective (commonly understood the same between human experts,) and word-domains of dataset attributes have their inherent semantic structures. It suggests that to limit the intuitive human influences on such construction, in this study, it requires that the constructed fuzzy set (fs-) representations of the declared word-sets should be the isomorphic images of their words. Such fs-representations of the word-domains are called, in this study, interpretable based on the concept of interpretability in the math-logical theories of A. Tarski et al. It requires the interpretability of the inherent semantic structures of the declared word-sets in their fs-representations structures. With this new feature, the study proposes a data-summarization method that can reveal L-distributions of fuzzy groups of objects represented by a given dataset to the desired dataset L-attribute. The set of all such mined LSs satisfies essential specific human usual L-knowledge, the scalability of its current attributes word-sets, and the current knowledge itself. An experimental study using the Bank Marketing dataset taken from the UCI dataset repository is performed to show the specific advantages of the proposed method.