Optimizing Convolutional Neural Network Parameters Using Genetic Algorithm for Breast Cancer Classification PDF Download
Are you looking for read ebook online? Search for your book and save it on your Kindle device, PC, phones or tablets. Download Optimizing Convolutional Neural Network Parameters Using Genetic Algorithm for Breast Cancer Classification PDF full book. Access full book title Optimizing Convolutional Neural Network Parameters Using Genetic Algorithm for Breast Cancer Classification by Khatereh Davoudi. Download full books in PDF and EPUB format.
Author: Khatereh Davoudi Publisher: ISBN: Category : Languages : en Pages : 0
Book Description
Breast cancer, as the most-regularly diagnosed cancer in women, can be controlled effectively by early-stage tumour diagnosis. Clinical specialists use Computer-Aided Diagnosis (CAD) systems to help aid in their diagnosis, as accurate as possible. Deep learning techniques, such as Convolutional Neural Network (CNN), due to their classification capabilities, have been widely adopted in CAD systems. The parameters of the network, including the weights of the convolution filters, and the weights of the fully connected layers play a crucial role in classification accuracy. Back-propagation technique is the most frequently used approach for training CNN. However, this technique has some disadvantages, such as getting stuck in local minima. In this thesis, we propose to optimize the weights of the CNN using Genetic Algorithm (GA). The work consists of: designing a CNN model to facilitate the classification process, training the model using three different optimizer (mini-batch gradient descent, Adam, and GA), and evaluating the model through various experiments on BreakHis dataset. We show that the CNN model trained through GA performs as well as the Adam optimizer with a classification accuracy of 85%.
Author: Khatereh Davoudi Publisher: ISBN: Category : Languages : en Pages : 0
Book Description
Breast cancer, as the most-regularly diagnosed cancer in women, can be controlled effectively by early-stage tumour diagnosis. Clinical specialists use Computer-Aided Diagnosis (CAD) systems to help aid in their diagnosis, as accurate as possible. Deep learning techniques, such as Convolutional Neural Network (CNN), due to their classification capabilities, have been widely adopted in CAD systems. The parameters of the network, including the weights of the convolution filters, and the weights of the fully connected layers play a crucial role in classification accuracy. Back-propagation technique is the most frequently used approach for training CNN. However, this technique has some disadvantages, such as getting stuck in local minima. In this thesis, we propose to optimize the weights of the CNN using Genetic Algorithm (GA). The work consists of: designing a CNN model to facilitate the classification process, training the model using three different optimizer (mini-batch gradient descent, Adam, and GA), and evaluating the model through various experiments on BreakHis dataset. We show that the CNN model trained through GA performs as well as the Adam optimizer with a classification accuracy of 85%.
Author: Pratheep Kumar Publisher: Infinite Study ISBN: Category : Mathematics Languages : en Pages : 11
Book Description
Decision tree algorithm is one of the algorithm which is easily understandable and interpretable algorithm used in both training and application purpose during breast cancer prognosis. To address this problem, Random Decision Forests are proposed. In this manuscript, the breast cancer classification can be determined by combining the advantages of Feature Weight and Hyper Parameter Tuned Random Decision Forest classifier
Author: Anthony Joseph Reiling Publisher: ISBN: Category : Genetic algorithms Languages : en Pages : 32
Book Description
This thesis proposes the use of a genetic algorithm (GA) to optimize the accuracy of a convolutional neural network (CNN). The GA modifies the structure of the CNN such as the number of convolutional filters, strides, kernel size, nodes, learning parameters, etc. Each modification of the network is trained and evaluated. Mutation of evolved networks create more successful networks over multiple generations. The final evolved network is 4.77% more accurate than a network proposed in the previous literature. Additionally, the evolved network is 13.4% less computationally complex.
Author: Ali Al Bataineh Publisher: ISBN: Category : Artificial immune systems Languages : en Pages : 0
Book Description
This dissertation presents efficient immunocomputing methods based on the clonal selection algorithms (CSA), a class of optimization procedures inspired by the clonal selection theory of adaptive immunity to optimize the learning ability of three main classes of artificial neural networks (ANNs): (1) multi-layer perceptrons (MLPs), (2) convolutional neural networks (CNNs), and (3) long short-term memory (LSTM)-based recurrent neural networks (RNNs). The first approach of this research is an application of CSA to train MLPs with predefined shallow architectures to optimize the model parameters (weights and biases). Gradient descent is the most widely used method to train MLPs because it is flexible and mathematically elegant. However, gradient descent requires the loss functions to be differentiable, and in some cases, it might converge to a set of sub-optimal weights and biases. Thus, this dissertation proposes a CSA-based approach as a competitive alternative to address the described problems. The performance of our proposed approach is compared with other training methods: genetic algorithm (GA), ant colony optimization (ACO), particle swarm optimization (PSO), and gradient descent with momentum. The comparison is benchmarked using five classification datasets: Iris Flower, Sonar, Wheat Seeds, Breast Cancer Wisconsin, and Haberman's Survival. The comparative study results show that CSA outperforms other training methods in all datasets; hence it can be employed as an effective method for training MLPs. The second proposed approach of our study is the application of CSA to discover the optimal hyperparameters of a deep CNN architecture in a fully automated manner to address image classification tasks without the necessity of expert knowledge. Most of the state-of-the-art CNN architectures are manually designed, which requires in-depth knowledge and can be time-consuming. In our second approach, the proposed methodology is evaluated on the EMNIST-Digits dataset. The results show that CSA can discover high-performance and less expensive CNN architectures in terms of the number of trainable parameters. Moreover, the EMNIST Digits dataset's optimized architecture is evaluated on other EMNIST datasets with increased data instances and classes. The results are impressive and demonstrate that the CSA finds efficient, reusable CNN architecture that can work for multiple datasets and still achieve competitive performance with the state-of-the-art. The third and final approach of the study proposes a CSA application to automatically design the architecture with optimal hyperparameters of the LSTM model for text classification tasks such as sentiment analysis and SMS spam classification. Similar to CNNs, designing LSTM's architectures requires expert domain knowledge and can be very time-consuming. The proposed methodology is evaluated on the large movie review dataset (IMDB). Furthermore, the architecture discovered by CSA for the IMDB dataset is also evaluated on the other datasets viz, Twitter US Airline Sentiment, and SMS Spam Collection. Additionally, the optimized LSTM architecture is combined with pre-determined CNN layers to achieve the same or better performance in less time and with fewer trainable parameters. For further verification and evaluation of the generalization ability and effectiveness of the proposed approach, it is compared with four machine learning algorithms widely used for text classification tasks: (1) random forest, (2) logistic regression, (3) support vector machine (SVM), (4) and multinomial naive Bayes. The results of our experiments show that the LSTM architecture automatically designed by our CSA method is more efficient, reusable and outperforms the machine learning algorithms and other models in the literature evaluated on the same three dataset. With the proposed CSA-based methods, the best CNN and LSTM architectures can be self-determined without any human intervention, making our CSA-based methods a promising approach to automatically discover optimal deep neural network architectures that provide the best performance for a given task.
Author: Janusz Kacprzyk Publisher: Springer Nature ISBN: 3031352483 Category : Technology & Engineering Languages : en Pages : 882
Book Description
This book describes the potential contributions of emerging technologies in different fields as well as the opportunities and challenges related to the integration of these technologies in the socio-economic sector. In this book, many latest technologies are addressed, particularly in the fields of computer science and engineering. The expected scientific papers covered state-of-the-art technologies, theoretical concepts, standards, product implementation, ongoing research projects, and innovative applications of Sustainable Development. This new technology highlights, the guiding principle of innovation for harnessing frontier technologies and taking full profit from the current technological revolution to reduce gaps that hold back truly inclusive and sustainable development. The fundamental and specific topics are Big Data Analytics, Wireless sensors, IoT, Geospatial technology, Engineering and Mechanization, Modeling Tools, Risk analytics, and preventive systems.
Author: Víctor Yepes Publisher: MDPI ISBN: 3039436074 Category : Technology & Engineering Languages : en Pages : 300
Book Description
In the current context of the electronic governance of society, both administrations and citizens are demanding the greater participation of all the actors involved in the decision-making process relative to the governance of society. This book presents collective works published in the recent Special Issue (SI) entitled “Optimization for Decision Making II”. These works give an appropriate response to the new challenges raised, the decision-making process can be done by applying different methods and tools, as well as using different objectives. In real-life problems, the formulation of decision-making problems and the application of optimization techniques to support decisions are particularly complex and a wide range of optimization techniques and methodologies are used to minimize risks, improve quality in making decisions or, in general, to solve problems. In addition, a sensitivity or robustness analysis should be done to validate/analyze the influence of uncertainty regarding decision-making. This book brings together a collection of inter-/multi-disciplinary works applied to the optimization of decision making in a coherent manner.
Author: Yuriy Zaychenko Publisher: Cambridge Scholars Publishing ISBN: 1527515400 Category : Medical Languages : en Pages : 133
Book Description
This book is devoted to the problems of information technologies (IT) and artificial intelligence methods applied to medical image processing, tumour detection and cancer classification in different human organs, including the breasts, lungs and brain. The most efficient modern tools in the problem of medical images processing and analysis are considered- convolutional neural networks (CNN). The main goal of this book is to present and analyze new perspective architectures of CNN aimed to increase accuracy of cancer classification. This book contains new approaches for improving efficiency of cancer detection in comparison with known CNN structures. The numerous experimental investigations proved their better efficiency by different classification criteria as compared with known. This book will be useful to specialists engaged in IT applications in medicine, dealing with development and application of medical diagnostics systems, students and postgraduates in Computer Science, all persons who are interested in IT applications in medicine, medical personnel engaged in malignant tumour diagnostics and cancer detection, and the wider public interested in the problems of cancer diagnostics that desire to extend their knowledge of prospective IT methods and their effectively solutions.
Author: Jason C. Hung Publisher: Springer Nature ISBN: 9819920922 Category : Computers Languages : en Pages : 1042
Book Description
This book comprises select peer-reviewed proceedings of the 6th International Conference on Innovative Computing (IC 2023). The contents focus on communication networks, business intelligence and knowledge management, web intelligence, and fields related to the development of information technology. The chapters include contributions on various topics such as databases and data mining, networking and communications, web and Internet of Things, embedded systems, soft computing, social network analysis, security and privacy, optical communication, and ubiquitous/pervasive computing. This volume will serve as a comprehensive overview of the latest advances in information technology for those working as researchers in both academia and industry.
Author: Nilanjan Dey Publisher: Academic Press ISBN: 012816087X Category : Science Languages : en Pages : 348
Book Description
Machine Learning in Bio-Signal Analysis and Diagnostic Imaging presents original research on the advanced analysis and classification techniques of biomedical signals and images that cover both supervised and unsupervised machine learning models, standards, algorithms, and their applications, along with the difficulties and challenges faced by healthcare professionals in analyzing biomedical signals and diagnostic images. These intelligent recommender systems are designed based on machine learning, soft computing, computer vision, artificial intelligence and data mining techniques. Classification and clustering techniques, such as PCA, SVM, techniques, Naive Bayes, Neural Network, Decision trees, and Association Rule Mining are among the approaches presented. The design of high accuracy decision support systems assists and eases the job of healthcare practitioners and suits a variety of applications. Integrating Machine Learning (ML) technology with human visual psychometrics helps to meet the demands of radiologists in improving the efficiency and quality of diagnosis in dealing with unique and complex diseases in real time by reducing human errors and allowing fast and rigorous analysis. The book's target audience includes professors and students in biomedical engineering and medical schools, researchers and engineers. Examines a variety of machine learning techniques applied to bio-signal analysis and diagnostic imaging Discusses various methods of using intelligent systems based on machine learning, soft computing, computer vision, artificial intelligence and data mining Covers the most recent research on machine learning in imaging analysis and includes applications to a number of domains
Author: Seyedali Mirjalili Publisher: Springer ISBN: 3319930257 Category : Technology & Engineering Languages : en Pages : 164
Book Description
This book introduces readers to the fundamentals of artificial neural networks, with a special emphasis on evolutionary algorithms. At first, the book offers a literature review of several well-regarded evolutionary algorithms, including particle swarm and ant colony optimization, genetic algorithms and biogeography-based optimization. It then proposes evolutionary version of several types of neural networks such as feed forward neural networks, radial basis function networks, as well as recurrent neural networks and multi-later perceptron. Most of the challenges that have to be addressed when training artificial neural networks using evolutionary algorithms are discussed in detail. The book also demonstrates the application of the proposed algorithms for several purposes such as classification, clustering, approximation, and prediction problems. It provides a tutorial on how to design, adapt, and evaluate artificial neural networks as well, and includes source codes for most of the proposed techniques as supplementary materials.
Author: Khatereh Davoudi
Author: Khatereh Davoudi
Author: Pratheep Kumar
Author: Anthony Joseph Reiling
Author: Ali Al Bataineh
Author: Janusz Kacprzyk
Author: Víctor Yepes
Author: Yuriy Zaychenko
Author: Jason C. Hung
Author: Nilanjan Dey
Author: Seyedali Mirjalili