Atoosa Madadkar, Masoud Karimlo, Mahdi Rahgozar, Hamid Jamaldini, Reza Mozaffari,
Volume 3, Issue 2 (10-2013)
Abstract
Objective: In the studies of genomics, it is essential to select a small number of single nucleotide polymorphisms (SNPs). That is more significant than the others for the association studies of disease susceptibility. Data mining technology provides an important means for extracting valuable medical rules hidden in medical data and acts as an important role in disease prediction and clinical diagnosis. In this study, our goal was to compare two machine learning methods using genetic factor and single nucleotide polymorphisms.
Methods: In order to perform the data analysis, a total of 141 patients and 83 controls in the genetics' section of Shahid Rajaee's heart center. The blood samples to draw conclusions about the LDLR and PCSK9 genes' SNPs was used. Also, the random forest and CART was used in order to discover the relationship between CAD and SNPs. These models were assessed by using four criteria including: sensitivity, specificity, precision and error. Data analysis was performed by SPSS (16.0) and R (2.15.0).
Results: CART had the better performance than Random Forest. Sensitivity, specificity, precision and error were 0.893, 0.506, 0.250 and 0.754 relatively. We introduced an algorithm to classify the high risk and low risk cases.
Conclusion:CART is suggested in order to assess the relationship between CAD and SNPs.
Seyyed Mohammad Mehdi Fatemi Bushehri, Mohsen Sardari Zarchi,
Volume 7, Issue 0 (4-2017)
Abstract
Abstract
Background & objective: The aim of this study is proposing an intelligent model for diagnosis and classification of learning disabilities based on machine learning methods and artificial neural network. Learning disabilities are among the most important and the most complex disabilities in the field of exceptional children's education. Exceptional education is an important area to which computer systems have contributed. Perhaps the first step in the education of exceptional children is the identification and classification of problems that these children face. A lot of research has been carried out regarding the use of machine learning techniques and artificial intelligence in the diagnosis and classification of learning disabilities. Reviewing of related works shows that machine learning techniques and expert systems are helpful to teachers and exceptional education’s specialists. Due to complex nature of and large number of learning disabilities, experts find it difficult to diagnose and classify learning disabilities without the help of computers. Insufficient number of experts raises work pressure and diagnosis delaying. Delaying in learning disabilities diagnosis causes various problems in learning disabilities treatment. The diversity and extent of learning disabilities and insufficient number of experts make an expert system necessary for the diagnosis and classification learning disabilities of children.
Methods: In this research firstly, the necessity to develop an expert system for classifying learning disabilities is discussed. Then with reviewing related works, strengths and weaknesses of each model is expressed. Digital signal processing, digital image processing and machine learning are the most cited methods used for learning disabilities classification in previous research. A review of the literature shows that models based on digital signal processing and digital image processing could not be used for this purpose because they are costly and require controlled conditions for analysis of digital signals and digital images. However, models based on digital signal processing and image processing are highly accurate. Models based on machine learning and artificial intelligence methods are also highly accurate. In addition, results show models based on artificial intelligence are less costly than models based on digital signal processing and image processing. Therefore, models based on machine learning methods are more appropriate than models based on digital signal processing and image processing for application systems. Artificial neural network could classify learning disabilities with an accuracy of over 85%.
Results: Results show that by using genetic algorithm for feature selection the accuracy of classification can be improved. In addition, by using fuzzy logic system researchers can extract rules of classification.
Conclusion: A hybrid intelligent model based on artificial intelligence and machine-learning methods using the strengths of previous models is proposed. The proposed model uses genetic algorithm for feature selection from among a set of features that have the highest impact in classification extracting. In the proposed model learning disabilities are classified with an artificial neural network. This model uses a fuzzy logic system to extracts rules of classification intelligently. The proposed model is highly accurate in classification and implementation simplicity. Finally, implementation of proposed model is explained.
