Impact of Unbalanced Classification on the Performance of Software Defect Prediction Models

Eldho, K. J.

doi:10.17485/ijst/v15i6.2193

Cited by 4 publications

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Performance of the classifiers is evaluated with the same metrics as that of ML-STAR: that is Precision (P), Recall (R) and fmeasure (F). The classifier results are discussed in two perspectives: first is analyzing the performance with varying number of instances with k-NN and Decision tree classifiers (12,13) . Second is obtaining the classifier performance at various pruning rules from no prune (R0) to rule 9 (R9).…”

Section: Results Of Classifiersmentioning

confidence: 99%

KNN Based Under Sampling: A Cognitive Centred Solution for Imbalanced Dataset Problem in Anaphora Resolution

Deepa¹,

Priya²,

Velvizhy³

2023

IJST

View full text Add to dashboard Cite

Background: Like many other real world applications, the machine learning system of anaphora resolution also struggles with skewed data. The problem of imbalanced classes occurs with classification task where there a huge difference exists in the number of instances among the involved classes.Objectives: The proposed framework intends to remove the imbalance first between positive and negative class instances before classifying them by KBUS that makes use of cognitive knowledge about the language and analysis is done at attribute level. Method: Nine pruning rules are crafted by KBUS(KNN Based Under Sampling) for TDIL dataset. Findings: During experimentation, number of positive instances are increased from 5.32% to 43.95%, whereas the number of negative instances are decreased from 94.68% to 56.05%. Loss ratio of positive and negative instance is 1:112. Finally the pruned dataset is classified by a list of classifiers namelyNaïve Bayes, SVM, Random forest, decision tree and k-NN. Novelty: Classifier results are discussed in two perspectives: Firstly the number of input instances and secondly the performance improvement achieved after pruning. It is adduced that pruning shows a remarkable improvement for all the classifiers. The proposed system produced an encouraging result as 78% of f-measure for k-NN and 77% for decision tree. Performance is presented in a comparative manner before and after pruning and the improvement of fmeasureranges from 13% (k-NN) to 41% (Random Forest). Thus this work has come up with a machine learning model to resolve Tamil anaphoric situations effectively in an imbalanced classification environment.

show abstract

Section: Results Of Classifiersmentioning

confidence: 99%

KNN Based Under Sampling: A Cognitive Centred Solution for Imbalanced Dataset Problem in Anaphora Resolution

Deepa¹,

Priya²,

Velvizhy³

2023

IJST

View full text Add to dashboard Cite

show abstract

“…The accuracy in detection was achieved up to 87%, whereas the classification is done, which is not impressive (19) . Other ML and DL models were introduced to detect multiple diseases in plants, which showed a progressive increase in terms of detection and classification accuracy (20)(21)(22)(23) . Though all prevailing approaches have not shown the remarkable performance, the new method is proposed in this research for efficient detection and classification of ALD.…”

Section: Introductionmentioning

confidence: 99%

Real Time Classification and Detection of Apple Leaf Diseases at Early Stage by Employing Enhanced Paddy Field Pattern Search Method (EPF-PSM) with 2LC

Donu,

Santhi

2023

IJST

View full text Add to dashboard Cite

Objectives: To propose a suitable machine learning approach to detect and classify the leaf diseases in apple trees at the pre-mature state to reduce plant degradation. In order to boost the accuracy level, EPF-PSM (Enhanced Paddy Field-Based Pattern Search Method) is employed along with 2LC (Two Level Classifier) technique. Methods: Preprocessing images, which include FS and FE, is done by employing EPF-PSM to detect the apple leaf disease. 2LC classifier is utilized to classify the type of disease based on the available features like gradient, pixel, edges, etc. The decision tree algorithm is employed to search for patterns in images and compare them for accurate prediction. With MC-Apriori (Multi Label Classification Apriori), the relationship between infected and non-infected leaves in search space is explored, and RGB scheme (RGBCS) is employed to discover the colour depth ratio of infected leaves. EPF-PSM extracts attributes from the ALDD-NW dataset (collected at the Apple Experiment Station of Northwest A&F University, China) and selects the suitable function to perform ALD classification, pre-mature detection, and accuracy. MATLAB software is utilized for implementation and assessment of the novel ML approach. The findings are compared with contemporary models such as the IFPA-GA with SVM-SVI, the FR-CNN, the R-SSD, and the INARSSD. Findings: The suggested model gives 98.01% accuracy rate, 97.06% detection speed in 29 seconds, 98.04% sensitivity, 95.92% specificity, 98.01% precision, 96.17% recall, 98.4% TPR, 99.03% TNR, and 96.07% F-Score, which is comparatively higher than the existing methods in terms of detection and classification of ALD. Novelty: The results shows that the suggested machine learning (ML) technique EPF-PSM with 2L-C has the ability to produce accurate detection and classification of ALD at the pre-mature stage, which helps the farmers to treat apple plants properly. The evident results outperform the prevailing methods IFPA-GA with SVM-SVI, FR-CNN, R-SSD, and INARSSD. https://www.indjst.org/

show abstract

“…INAR-SSD is used to detect discriminant features in plant leaves automatically and detect disease at an early stage by combining algorithms such as rainbow concatenation and single-and multi-scale object detection methods (22). The author discusses imbalanced classification models in software defect prediction, where the object detection and accuracy levels are thoroughly tested without any flaws (23). Some of the methods comply to detect the apple leaf diseases in early stage is tested and it is mentioned below.…”

Section: Introductionmentioning

confidence: 99%

Early Detection and Classification of Apple Leaf Diseases by utilizing IFPA Genetic Algorithm with MC-SVM, SVI and Deep Learning Methods

Jose¹,

Santhi²

2022

IJST

View full text Add to dashboard Cite

Objectives:To propose a new model for early detection and classification of apple leaf diseases by means of genetic algorithms and advanced deep learning methods in order to minimize plant degradation and maximize detection accuracy. Methods: A feature selection and extraction task is carried out by using an Improvised Flower Pollination Technique (IFPA) based genetic algorithm. Deep learning techniques such as Multi Class Support Vector Machine (MC-SVM) and Spectral Vegetation Indices (SVI) are used to classify and detect disease. The Region Growing Algorithm (RGA) and the Stochastic Gradient Descent Method (SGDM) are used to detect leaf colour segments and identify the disease at an early stage based on shape, texture, and colour features. IFPA-GA with MC-SVM and SVI extracts the features from the apple leaf dataset (collected from the Apple Experiment Station of Northwest A&F University, China) and then selects the appropriate function to perform classification and detection of disease at early occurrence and to improve the accuracy level. To demonstrate the efficiency of the proposed algorithm, MATLAB is used for implementation. The performance results are evaluated and compared to the existing models such as Faster R-CNN, R-SSD, and INAR-SSD. Findings: Early disease detection and classification of leaf diseases are achieved with 94.09% accuracy level, 93.07% Speed, 94.01% sensitivity, 93.38% specificity, 95. 01% precision, 93.17% recall, 190 True Positive, 110 True Negative, and 92.07% F-Score to detect the disease and classify it in an optimized manner, which is high compared to the existing versions. Novelty: According to the findings of the comprehensive study, the proposed detection and classification method IFPA-GA with SVM-SVI outperforms Faster R-CNN, R-SSD, and INAR-SSD in terms of accuracy and speed of apple leaf disease detection at an early occurrence by classifying it in a robust manner.

show abstract

Impact of Unbalanced Classification on the Performance of Software Defect Prediction Models

Cited by 4 publications

References 18 publications

KNN Based Under Sampling: A Cognitive Centred Solution for Imbalanced Dataset Problem in Anaphora Resolution

KNN Based Under Sampling: A Cognitive Centred Solution for Imbalanced Dataset Problem in Anaphora Resolution

Real Time Classification and Detection of Apple Leaf Diseases at Early Stage by Employing Enhanced Paddy Field Pattern Search Method (EPF-PSM) with 2LC

Early Detection and Classification of Apple Leaf Diseases by utilizing IFPA Genetic Algorithm with MC-SVM, SVI and Deep Learning Methods

Contact Info

Product

Resources

About