Feature selection with decision tree criterion

Accurate and fast characterization of the subtype sequences of Avian influenza A virus (AIAV) hemagglutinin (HA) and neuraminidase (NA) depends on expanding diagnostic services and is embedded in molecular epidemiological studies. A new approach for classifying the AIAV sequences of the HA and NA genes into subtypes using DNA sequence data and physicochemical properties is proposed. This method simply requires unaligned, full-length, or partial sequences of HA or NA DNA as input. It allows for quick and highly accurate assignments of HA sequences to subtypes H1–H16 and NA sequences to subtypes N1–N9. For feature extraction, k-gram, discrete wavelet transformation, and multivariate mutual information were used, and different classifiers were trained for prediction. Four different classifiers, Naïve Bayes, Support Vector Machine (SVM), K nearest neighbor (KNN), and Decision Tree, were compared using our feature selection method. This comparison is based on the 30% dataset separated from the original dataset for testing purposes. Among the four classifiers, Decision Tree was the best, and Precision, Recall, F1 score, and Accuracy were 0.9514, 0.9535, 0.9524, and 0.9571, respectively. Decision Tree had considerable improvements over the other three classifiers using our method. Results show that the proposed feature selection method, when trained with a Decision Tree classifier, gives the best results for accurate prediction of the AIAV subtype.

Download Full-text

A Contemporary Machine Learning Method for Accurate Prediction of Cervical Cancer

SHS Web of Conferences ◽

10.1051/shsconf/202110204004 ◽

2021 ◽

Vol 102 ◽

pp. 04004

Author(s):

Jesse Jeremiah Tanimu ◽

Mohamed Hamada ◽

Mohammed Hassan ◽

Saratu Yusuf Ilu

Keyword(s):

Machine Learning ◽

Cervical Cancer ◽

Feature Selection ◽

Decision Tree ◽

Sensitivity And Specificity ◽

Missing Values ◽

New Technologies ◽

Machine Learning Techniques ◽

Screening Tests ◽

Tree Classifier

With the advent of new technologies in the medical field, huge amounts of cancerous data have been collected and are readily accessible to the medical research community. Over the years, researchers have employed advanced data mining and machine learning techniques to develop better models that can analyze datasets to extract the conceived patterns, ideas, and hidden knowledge. The mined information can be used as a support in decision making for diagnostic processes. These techniques, while being able to predict future outcomes of certain diseases effectively, can discover and identify patterns and relationships between them from complex datasets. In this research, a predictive model for predicting the outcome of patients’ cervical cancer results has been developed, given risk patterns from individual medical records and preliminary screening tests. This work presents a Decision tree (DT) classification algorithm and shows the advantage of feature selection approaches in the prediction of cervical cancer using recursive feature elimination technique for dimensionality reduction for improving the accuracy, sensitivity, and specificity of the model. The dataset employed here suffers from missing values and is highly imbalanced. Therefore, a combination of under and oversampling techniques called SMOTETomek was employed. A comparative analysis of the proposed model has been performed to show the effectiveness of feature selection and class imbalance based on the classifier’s accuracy, sensitivity, and specificity. The DT with the selected features and SMOTETomek has better results with an accuracy of 98%, sensitivity of 100%, and specificity of 97%. Decision Tree classifier is shown to have excellent performance in handling classification assignment when the features are reduced, and the problem of imbalance class is addressed.

Download Full-text

PCA based Regression Decision Tree Classification for Somatic Mutations.

International Journal of Engineering and Advanced Technology - Regular Issue ◽

10.35940/ijeat.f1181.0986s319 ◽

2019 ◽

Vol 8 (6S3) ◽

pp. 1095-1102

Keyword(s):

Feature Selection ◽

Decision Tree ◽

Somatic Mutations ◽

Feature Selection Technique ◽

Data Set ◽

Cancer Data ◽

Classification Feature ◽

Decision Tree Classification ◽

Proposed Model ◽

Feature Selection Techniques

The analization of cancer data and normal data for the predication of somatic mu-tation occurrences in the data set plays an important role and several challenges persist in detectingsomatic mutations which leads to complexity of handling large volumes of data in classifi-cation with good accuracy. In many situations the dataset may consist of redundant and less significant features and there is a need to remove insignificant features in order to improve the performance of classification. Feature selection techniques are useful for dimensionality reduction purpose. PCA is one type of feature selection technique to identify significant attributes and is adopted in this paper. A novel technique, PCA based regression decision tree is proposed for classification of somatic mutations data in this paper.The performance analysis of this clas-sification process for the detection of somatic mutation is compared with existing algorithms and satisfactory results are obtained with the proposed model.

Download Full-text

A DECISION TREE-BASED CLASSIFICATION FRAMEWORK FOR USED OIL ANALYSIS APPLYING RANDOM FOREST FEATURE SELECTION

Journal of Applied Science, Engineering and Technology for Development ◽

10.33803/jasetd.2017.3-1.7 ◽

2018 ◽

Keyword(s):

Decision Making ◽

Feature Selection ◽

Random Forest ◽

Decision Tree ◽

Condition Monitoring ◽

Critical Parameters ◽

Oil Analysis ◽

Used Oil ◽

Maintenance Decision

Lubricant condition monitoring (LCM), part of condition monitoring techniques under Condition Based Maintenance, monitors the condition and state of the lubricant which reveal the condition and state of the equipment. LCM has proved and evidenced to represent a key concept driving maintenance decision making involving sizeable number of parameter (variables) tests requiring classification and interpretation based on the lubricant’s condition. Reduction of the variables to a manageable and admissible level and utilization for prediction is key to ensuring optimization of equipment performance and lubricant condition. This study advances a methodology on feature selection and predictive modelling of in-service oil analysis data to assist in maintenance decision making of critical equipment. Proposed methodology includes data pre-processing involving cleaning, expert assessment and standardization due to the different measurement scales. Limits provided by the Original Equipment Manufacturers (OEM) are used by the analysts to manually classify and indicate samples with significant lubricant deterioration. In the last part of the methodology, Random Forest (RF) is used as a feature selection tool and a Decision Tree-based (DT) classification of the in-service oil samples. A case study of a thermal power plant is advanced, to which the framework is applied. The selection of admissible variables using Random Forest exposes critical used oil analysis (UOA) variables indicative of lubricant/machine degradation, while DT model, besides predicting the classification of samples, offers visual interpretability of parametric impact to the classification outcome. The model evaluation returned acceptable predictive, while the framework renders speedy classification with insights for maintenance decision making, thus ensuring timely interventions. Moreover, the framework highlights critical and relevant oil analysis parameters that are indicative of lubricant degradation; hence, by addressing such critical parameters, organizations can better enhance the reliability of their critical operable equipment.

Download Full-text

A Comparison of the Analysis of Methods for Feature Extraction and Classification by Wavelet Transform in SSVEP BCIs

10.21203/rs.3.rs-82008/v1 ◽

2020 ◽

Author(s):

Hoda Heidari ◽

Zahra Einalou ◽

Mehrdad Dadgostar ◽

Hamidreza Hosseinzadeh

Keyword(s):

Feature Extraction ◽

Feature Selection ◽

Wavelet Transform ◽

Decision Tree ◽

Nearest Neighbor ◽

Support Vector ◽

K Nearest Neighbor ◽

Iir Filters ◽

Wide Range ◽

New Feature

Abstract Most of the studies in the field of Brain-Computer Interface (BCI) based on electroencephalography have a wide range of applications. Extracting Steady State Visual Evoked Potential (SSVEP) is regarded as one of the most useful tools in BCI systems. In this study, different methods such as feature extraction with different spectral methods (Shannon entropy, skewness, kurtosis, mean, variance) (bank of filters, narrow-bank IIR filters, and wavelet transform magnitude), feature selection performed by various methods (decision tree, principle component analysis (PCA), t-test, Wilcoxon, Receiver operating characteristic (ROC)), and classification step applying k nearest neighbor (k-NN), perceptron, support vector machines (SVM), Bayesian, multiple layer perceptron (MLP) were compared from the whole stream of signal processing. Through combining such methods, the effective overview of the study indicated the accuracy of classical methods. In addition, the present study relied on a rather new feature selection described by decision tree and PCA, which is used for the BCI-SSVEP systems. Finally, the obtained accuracies were calculated based on the four recorded frequencies representing four directions including right, left, up, and down.

Download Full-text

ANGEL Mining

Higher Education Institutions and Learning Management Systems ◽

10.4018/978-1-60960-884-2.ch005 ◽

2012 ◽

pp. 94-115 ◽

Cited By ~ 2

Author(s):

Tyler Swanger ◽

Kaitlyn Whitlock ◽

Anthony Scime ◽

Brendan P. Post

Keyword(s):

Data Mining ◽

Decision Making ◽

Feature Selection ◽

Decision Tree ◽

Management System ◽

Association Mining ◽

Data Mining Techniques ◽

Usage Patterns ◽

Design Selection ◽

Comprehensive College

This chapter data mines the usage patterns of the ANGEL Learning Management System (LMS) at a comprehensive college. The data includes counts of all the features ANGEL offers its users for the Fall and Spring semesters of the academic years beginning in 2007 and 2008. Data mining techniques are applied to evaluate which LMS features are used most commonly and most effectively by instructors and students. Classification produces a decision tree which predicts the courses that will use the ANGEL system based on course specific attributes. The dataset undergoes association mining to discover the usage of one feature’s effect on the usage of another set of features. Finally, clustering the data identifies messages and files as the features most commonly used. These results can be used by this institution, as well as similar institutions, for decision making concerning feature selection and overall usefulness of LMS design, selection and implementation.

Download Full-text