Performance Analysis of State-of-the-Art CNN Architectures for LUNA16

Abstract

The convolutional neural network (CNN) has become a powerful tool in machine learning (ML) that is used to solve complex problems such as image recognition, natural language processing, and video analysis. Notably, the idea of exploring convolutional neural network architecture has gained substantial attention as well as popularity. This study focuses on the intrinsic various CNN architectures: LeNet, AlexNet, VGG16, ResNet-50, and Inception-V1, which have been scrutinized and compared with each other for the detection of lung cancer using publicly available LUNA16 datasets. Furthermore, multiple performance optimizers: root mean square propagation (RMSProp), adaptive moment estimation (Adam), and stochastic gradient descent (SGD), were applied for this comparative study. The performances of the three CNN architectures were measured for accuracy, specificity, sensitivity, positive predictive value, false omission rate, negative predictive value, and F1 score. The experimental results showed that the CNN AlexNet architecture with the SGD optimizer achieved the highest validation accuracy for CT lung cancer with an accuracy of 97.42%, misclassification rate of 2.58%, 97.58% sensitivity, 97.25% specificity, 97.58% positive predictive value, 97.25% negative predictive value, false omission rate of 2.75%, and F1 score of 97.58%. AlexNet with the SGD optimizer was the best and outperformed compared to the other state-of-the-art CNN architectures.

Performance analysis of state‐of‐the‐art CNN architectures for brain tumour detection

“…Where η: Initial learning rate; q t : Exponential average of gradient along w j p t : Gradient at time t along with w j s t : Exponential average of squares of gradient along w j γ 1 and γ 2 : Hyperparameter SGD (Stochastic gradient descent) is the third optimiser that we chose for performance evaluation, this optimiser updates the parameter one at a time. 3 Therefore, it is much faster than other optimisers and cost function minimised after each iteration. It requires less memory consumption and Due to its capability of frequently updating model parameter cost function can be fluctuating heavily, that why gradient jump to the global minimum point.…”

“…2 The domains of image analysis and computer vision are particularly well-suited for the application of deep learning to real-world situations. 3 When it comes to image classification and detection issues, the convolutional neural network (CNN) stands out as a top deep learning model, 4 due to its ability to automatically extract features from images and accurately classify or segment them. Medical imaging is currently the best method for identifying and treating diseases in their earliest stages.…”

“…The use of deep learning, a type of AI, has been shown to be effective in the detection of cytopathic abnormalities 2 . The domains of image analysis and computer vision are particularly well‐suited for the application of deep learning to real‐world situations 3 . When it comes to image classification and detection issues, the convolutional neural network (CNN) stands out as a top deep learning model, 4 due to its ability to automatically extract features from images and accurately classify or segment them.…”

Integrating of Image Processing and Number Recognition in Sudoku Puzzle Cards Digitation

“…The "■" feature appears in the upper-left corner, the lower-right corner, and the middle of the image. After calculations are performed for the convolution layer and pooling layer,  is similarly derived, meaning that the feature can be detected wherever it appears [24].…”

A Comprehensive Investigation into the Application of Convolutional Neural Networks (ConvNet/CNN) in Smart Grids

“…This is calculated for both x and ŷ. [69]. Another adaptive learning technique for addressing damaging learning rates is the root mean square propagation.…”