A Hierarchical Classifier for Multiclass Prostate Histopathology Image Gleason Grading

Albashish, Dheeb; Sahran, Shahnorbanun; Abdullah, Azizi; Alweshah, Mohammed; Adam, Afzan

doi:10.32890/jict2018.17.2.7

Cited by 6 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Support Vector Machine (SVM) is a supervised learning algorithm developed by Vapnik and others at AT and T Bell Laboratories (Cortes and Vapnik, 2004). SVM is one of popular machine learning algorithms for classification and has been extensively used with excellent empirical performance in computer security (Ariff et al, 2018;Zolfi et al, 2019;Kadis and Abdullah, 2017) and other research areas such as in computer vision (Albashish et al, 2018;Abdullah et al, 2009;Nashat et al, 2011). The method is intended for binary classification problems.…”

Section: Support Vector Machinesmentioning

confidence: 99%

“…However, one possible challenge in this approach is selecting parent (root) and child nodes of the tree. Furthermore, the selected nodes in the tree-based structure should be discriminative by considering the domain knowledge and the relationship between different classes of the base learners (Albashish et al, 2018;Zhang et al, 2018). Thus, in this study, we introduce a soft marking scheme to assess a set of binary classifiers to ensure the best overall predictive base learners to model the relationships of different attacks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Soft Marking Scheme of SVM Hierarchical Classifiers for Attack Classification

Abdullah¹

2021

Journal of Computer Science

Self Cite

View full text Add to dashboard Cite

Classification is a predictive modelling problem that involves assigning a class label to an instance correctly. There exist several strategies in machine learning to deal with the multi-class classification problems for attack detection. One of the popular strategies is the one-vs-one that decomposes the multi-class problem into multiple binary ones. The approach has been applied in many popular supervised learning algorithms, such as support vector machines. A possible problem of the standard multi-class classification problem is that it lacks correlation between different classes, which can increase overfitting problems and hinder generalization performance. Thus, a possible solution to the problem is to use a hierarchical classification that captures the relationship between classes by dividing the multi-class classification problem into a tree. However, one possible challenge in this approach is selecting parent and child nodes of the tree. The selected nodes should be informative to recognize and then classify different attack classes. One way is by looking at specific domain knowledge to train and build classifiers of the base learners for effective prediction. Thus, a soft marking scheme is introduced to assess a set of binary classifiers to ensure the best overall predictive base learners. Finally, we validate and compare the proposed approach to the standard NSL-KDD dataset. The results show that the proposed method outperforms the standard classifier on the intrusion attack classification.

show abstract

Section: Support Vector Machinesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Soft Marking Scheme of SVM Hierarchical Classifiers for Attack Classification

Abdullah¹

2021

Journal of Computer Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…This model can then be utilized to classify the class values of other instances. Examples of classification tasks include intrusion detection, medical diagnosis, handwritten digit recognition, spam email detection, and bankruptcy determination [5][6][7][8][9]. In recent years, the field of swarm intelligence (SI) has been derived by observing the swarming behavior of some animals, such as ant colonies, flocking birds, and fish schools [10,11].…”

Section: Introductionmentioning

confidence: 99%

Annealing strategy for an enhance rule pruning technique in ACO-Based rule classification

Al-Behadili

Ku-Mahamud

Sagban

2019

IJEECS

View full text Add to dashboard Cite

Ant colony optimization (ACO) was successfully applied to data mining classification task through ant-mining algorithms. Exploration and exploitation are search strategies that guide the learning process of a classification model and generate a list of rules. Exploitation refers to the process of intensifying the search for neighbors in good regions, whereas exploration aims towards new promising regions during a search process. The existing balance between exploration and exploitation in the rule construction procedure is limited to the roulette wheel selection mechanism, which complicates rule generation. Thus, low-coverage complex rules with irrelevant terms will be generated. This work proposes an enhancement rule pruning procedure for the ACO algorithm that can be used in rule-based classification. This procedure, called the annealing strategy, is an improvement of ant-mining algorithms in the rule construction procedure. Presented as a pre-pruning technique, the annealing strategy deals first with irrelevant terms before creating a complete rule through an annealing schedule. The proposed improvement was tested through benchmarking experiments, and results were compared with those of four of the most related ant-mining algorithms, namely, Ant-Miner, CAnt-Miner, TACO-Miner, and Ant-Miner with hybrid pruner. Results display that our proposed technique achieves better performance in terms of classification accuracy, model size, and computational time. The proposed annealing schedule can be used in other ACO variants for different applications to improve classification accuracy.

show abstract

“…In clinical practice, experts are required to perform the blood smears in response to a clinical feature or to a previously abnormal complete blood count in patients. They also have to manually classify the clump cells which is tedious, time-consuming and involves qualitative process [1], [2]. In addition, the existing methods contribute to inaccuracy, inconsistency and poor reliability diagnosis that may lead to false diagnosis situation.…”

Section: Introductionmentioning

confidence: 99%

Concavity Point and Skeleton Analysis Algorithm for Detection and Quantization in Heavily Clumped Red Blood Cells

Ahmad

Abdullah

Sabudin

2019

Int. J. Adv. Sci. Eng. Inf. Technol.

View full text Add to dashboard Cite

In practice, most hospitals use light microscope to examine the smeared blood for blood quantification. This visual quantification is subjective, laborious and time-consuming. Although automating the process is a good solution, the available techniques are unable to count or ignore the clumpy red blood cells (RBC). Moreover, clumping cell can affect the whole counting process of RBC as well as their accuracy. This paper proposes a new quantization process called concavity point and skeleton analysis (CP-SA) for heavily clump RBC. The proposed methodology is based on induction approach, enhanced lime blood cell by using gamma correction to get the appropriate edges. Then, splitting the clump and single cells by calculating each object area in pixel. Later, the quantification of clumpy cells with the proposed CP-SA method is done. This algorithm has been tested on 556 clump RBC taken from thin blood smear images under light microscope. All dataset images are captured from Hematology Unit, UKM Medical Centre in Kuala Lumpur. On all tested images, the cells of interest are successfully detected and counted from those clump cells. A comparative study and analysis to evaluate the performance of the proposed algorithm in three levels of clump have been conducted. The first level was with two clumps, second level with three clumps and third level with four clumps. The counting number of clump cells has been analyzed using quantitative analysis, resulting in much better results compared to other recent algorithms. The comparison shows that the proposed method gives better precision result at all levels with respect to ground truth: two clump cells (92%), three clump cells (96%) and four clump cells (90%). The results prove that this study has successfully developed a new method to count heavily clump RBC more accurately in microscopic images. In addition, this can be considered as a low-cost solution for quantification in massive examination.

show abstract

A Hierarchical Classifier for Multiclass Prostate Histopathology Image Gleason Grading

Cited by 6 publications

References 26 publications

Soft Marking Scheme of SVM Hierarchical Classifiers for Attack Classification

Soft Marking Scheme of SVM Hierarchical Classifiers for Attack Classification

Annealing strategy for an enhance rule pruning technique in ACO-Based rule classification

Concavity Point and Skeleton Analysis Algorithm for Detection and Quantization in Heavily Clumped Red Blood Cells

Contact Info

Product

Resources

About