Image Classification Based on the Boost Convolutional Neural Network

Lee, Shin-Jye; Chen, Tonglin; Yu, Li; Lai, Chin-Hui

doi:10.1109/access.2018.2796722

Cited by 73 publications

(64 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The The second part is the initiation of the function, where the obtained activation value is used for the nonlinear compressed transformation to extricate a nonlinear eigenvalue. The frequentlyused activation functions include ReLU, Sigmoid, and Tanh (Lee, Chen, Yu, & Lai, 2018). A neural network is a network based on the interconnection between artificial neurons.…”

Section: Multi-layer Perceptronmentioning

confidence: 99%

“…The feedforward neural network (FNN) or multilayer perceptron (MLP) is a neural network that permits the feedforward connection of neurons. The input of data is known as the input layer, while the output of results is termed as the output layer; the layers between the input layer and the output layer are called the hidden layers (Lee, Chen, Yu, & Lai, 2018). MLP is a supervised algorithm that learns a function (.…”

Section: Multi-layer Perceptronmentioning

confidence: 99%

“…(ordinal number) neuron on layer I and "j" (ordinal number) neural on layer I+1; denotes the connection between the bias on layer I and "j" neuron on layer I+1; and implies the activation value (output value) of the "I" neuron on layer I, and the activation value of the blue neuron in the picture is 2 (2) (Lee, Chen, Yu, & Lai, 2018).…”

Section: Multi-layer Perceptronmentioning

confidence: 99%

“…While the constant is set to ''1'' as the input of 0 th dimension, the bias is managed as the weight of 0 th dimension. This is also called affine transformation(Lee, Chen, Yu, & Lai, 2018).…”

mentioning

confidence: 99%

See 3 more Smart Citations

Prognosticating Mesothelioma Using Predictive Analytics

Choudhury¹

2019

Preprint

View full text Add to dashboard Cite

Background: Malignant pleural mesothelioma (MPM) is an atypical, belligerent tumor that matures into cancer in the pleura, a stratum of tissue bordering the lungs. Pleural mesothelioma is a common type of mesothelioma that accounts for about 75 percent of all mesothelioma diagnosed yearly in the United States. Diagnosis of mesothelioma takes several months and is expensive. Given the difficulty of diagnosing MPM, early identification is crucial for patient survival. Our study implements artificial intelligence and recommends the best fit model for early diagnosis and prognosis of MPM. Method: We retrospectively retrieved patient’s medical reports generated by Dicle University, Turkey and implemented multi-layered perceptron (MLP), voted perceptron (VP), Clojure classifier (CC), kernel logistic regression (KLR), stochastic gradient decent SGD), adaptive boosting (AdaBoost), Hoeffding tree (VFDT), and primal estimated sub-gradient solver for support vector machine (s-Pegasos). We evaluated the models, compared and tested using paired T-test (corrected) at 0.05 significance based on their respective classification accuracy, f-measure, precision, recall, root mean squared error, receivers characteristic curve (ROC), and precision-recall curve (PRC). Results: In phase-1 SGD, AdaBoost.M1, KLR, MLP, VFDT generates optimal results with the highest possible performance measures. In phase-2, AdaBoost with a classification accuracy of 71.29% outperformed all other algorithms. C-reactive protein, platelet count, duration of symptoms, gender, and pleural protein were found to be the most relevant predictors that can prognosticate mesothelioma. Conclusion: This study confirms that data obtained from biopsy and imagining tests are strong predictors of mesothelioma but are associated with high cost, however, can identify mesothelioma with optimal accuracy. Predictive analytics without using biopsy results can diagnose mesothelioma with acceptable accuracy. Implementation of phase-2 followed by phase-1 can address diagnosis expenses and maximize disease prognosis. Additionally, results indicate improved MPM diagnosis using AI methods dependent upon the specific application.

show abstract

Section: Multi-layer Perceptronmentioning

confidence: 99%

Section: Multi-layer Perceptronmentioning

confidence: 99%

Section: Multi-layer Perceptronmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Prognosticating Mesothelioma Using Predictive Analytics

Choudhury¹

2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Convolutional Neural Network (CNN) is an alternative and recent trend for image classification that has been proven to produce high accuracy in image classification tasks [17] without requiring any task-specific feature engineering [18]. It is considered the most successful machine learning model in recent years [19] and the most eminent method in computer vision [20], in part because it consists of a powerful image features extractor [21].…”

Section: Convolutional Neural Networkmentioning

confidence: 99%

Object Classification in Thermal Images using Convolutional Neural Networks for Search and Rescue Missions with Unmanned Aerial Systems

Rodin

Lima

Andrade

et al. 2018

2018 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

In recent years, the use of Unmanned Aerial Systems (UAS) has become commonplace in a wide variety of tasks due to their relatively low cost and ease of operation. In this paper, we explore the use of UAS in maritime Search And Rescue (SAR) missions by using experimental data to detect and classify objects at the sea surface. The objects are chosen as common objects present in maritime SAR missions: a boat, a pallet, a human, and a buoy. The data consists of thermal images and Gaussian Mixture Model (GMM) is used to discriminate foreground objects from the background. Then, bounding boxes containing the object are defined and used to train a Convolutional Neural Network (CNN). The CNN achieves the average accuracy of 92.5% when evaluating a testing dataset.

show abstract