Neural Deep Learning Research Articles

A deep neural network based classifier for brain tumor diagnosis

Classification process plays a key role in diagnosing brain tumors. Earlier research works are intended for identifying brain tumors using different classification techniques. However, the False Alarm Rates (FARs) of existing classification techniques are high. To improve the early-stage brain tumor diagnosis via classification the Weighted Correlation Feature Selection Based Iterative Bayesian Multivariate Deep Neural Learning (WCFS-IBMDNL) technique is proposed in this work. The WCFS-IBMDNL algorithm considers medical dataset for classifying the brain tumor diagnosis at an early stage. At first, the WCFS-IBMDNL technique performs Weighted Correlation-Based Feature Selection (WC-FS) by selecting subsets of medical features that are relevant for classification of brain tumors. After completing the feature selection process, the WCFS-IBMDNL technique uses Iterative Bayesian Multivariate Deep Neural Network (IBMDNN) classifier for reducing the misclassification error rate of brain tumor identification. The WCFS-IBMDNL technique was evaluated in JAVA language using Disease Diagnosis Rate (DDR), Disease Diagnosis Time (DDT), and FAR parameter through the epileptic seizure recognition dataset.

Learning to detect Android malware via opcode sequences

A large number of Android malware samples can be deployed as the variants of the previously known samples. In consequence, a classification system capable of supporting a large set of samples is required to secure Android platform. Although a large set of variants requires scalability for automatic detection and classification, it also presents a significant advantage about a richer dataset at the stage of discovering underlying malicious activities and extracting representative features. Deep Neural Networks are built by a complex structure of layers whose parameters can be tuned and trained in order to enhance classification statistical metric results. Emerging parallelization computing tools and processors reduce computation time.In this paper, we propose a deep learning Android malware detection method using features extracted from instruction call graphs. The presented method examines all possible execution paths and the balanced dataset improves deep neural learning benign execution paths versus malicious paths. Since there is not a publicly available model for Android malware detection, we train deep networks from scratch. Then, we apply a grid search method to seek the optimal parameters of the network and to discover the combination of the hyper-parameters, which maximizes the statistical metric values. To validate the effectiveness of the proposed method, we evaluate with a balanced dataset constituted by 24,650 malicious and 25,000 benign samples. We evaluate the deep network architecture with respect to different parameters and compare the statistical metric values including runtime with respect to baseline classifiers. Our experimental results show that the presented malware detection is reached at 91.42% level in accuracy and 91.91% in F-measure, respectively.

Extreme Learning Machine with Elastic Net Regularization

Compared with deep neural learning, the extreme learning machine (ELM) can be quickly converged without iteratively tuning hidden nodes. Inspired by this merit, an extreme learning machine with elastic net regularization (ELM-EN) is proposed in this paper. The elastic net is a regularization method that combines LASSO and ridge penalties. This regularization can keep a balance between system stability and solution's sparsity. Moreover, an excellent optimization method, i.e., accelerated proximal gradient, is used to find the minimum of the system optimization function. Various datasets from UCI repository and two facial expression image datasets are used to validate the efficiency of our system. Final experimental results indicate that our ELM-EN requires less training time than multi-layer perceptron, and can achieve higher recognition accuracy than ELM and sparse ELM.

Comparative study of singing voice detection based on deep neural networks and ensemble learning

This paper investigates various structures of neural network models and various types of stacked ensembles for singing voice detection. The studied models include convolutional neural networks (CNN), long short term memory (LSTM) model, convolutional LSTM model, and capsule net. The input features to the network models are MFCC (mel-frequency cepstrum coefficients), spectrogram from short-time Fourier transformation, or raw PCM samples. The simulation results show that CNN model with spectrogram inputs yields higher detection accuracy, up to 91.8% for Jamendo dataset. Among the studied stacked ensemble methods, performing voting strategy yields comparable performance as the other methods, but with much lower computational cost. By voting with five models, the accuracy reaches 94.2% for Jamendo dataset.

Prediction of response and toxicity to immune checkpoint inhibitor therapies (ICI) in melanoma using deep neural networks machine learning.

9529Background: Challenges in treating melanoma patients with ICI include treatment resistance and adverse events, both of which can lead to discontinuation of treatment. Histopathology of metastat...

White blood cells identification system based on convolutional deep neural learning networks

Background and objectivesWhite blood cells (WBCs) differential counting yields valued information about human health and disease. The current developed automated cell morphology equipments perform differential count which is based on blood smear image analysis. Previous identification systems for WBCs consist of successive dependent stages; pre-processing, segmentation, feature extraction, feature selection, and classification. There is a real need to employ deep learning methodologies so that the performance of previous WBCs identification systems can be increased. Classifying small limited datasets through deep learning systems is a major challenge and should be investigated. MethodsIn this paper, we propose a novel identification system for WBCs based on deep convolutional neural networks. Two methodologies based on transfer learning are followed: transfer learning based on deep activation features and fine-tuning of existed deep networks. Deep acrivation featues are extracted from several pre-trained networks and employed in a traditional identification system. Moreover, a novel end-to-end convolutional deep architecture called “WBCsNet” is proposed and built from scratch. Finally, a limited balanced WBCs dataset classification is performed through the WBCsNet as a pre-trained network. ResultsDuring our experiments, three different public WBCs datasets (2551 images) have been used which contain 5 healthy WBCs types. The overall system accuracy achieved by the proposed WBCsNet is (96.1%) which is more than different transfer learning approaches or even the previous traditional identification system. We also present features visualization for the WBCsNet activation which reflects higher response than the pre-trained activated one. Conclusiona novel WBCs identification system based on deep learning theory is proposed and a high performance WBCsNet can be employed as a pre-trained network.

Obtaining Cross Modal Similarity Metric with Deep Neural Architecture

Analyzing complex system with multimodal data, such as image and text, has recently received tremendous attention. Modeling the relationship between different modalities is the key to address this problem. Motivated by recent successful applications of deep neural learning in unimodal data, in this paper, we propose a computational deep neural architecture, bimodal deep architecture (BDA) for measuring the similarity between different modalities. Our proposed BDA architecture has three closely related consecutive components. For image and text modalities, the first component can be constructed using some popular feature extraction methods in their individual modalities. The second component has two types of stacked restricted Boltzmann machines (RBMs). Specifically, for image modality a binary-binary RBM is stacked over a Gaussian-binary RBM; for text modality a binary-binary RBM is stacked over a replicated softmax RBM. In the third component, we come up with a variant autoencoder with a predefined loss function for discriminatively learning the regularity between different modalities. We show experimentally the effectiveness of our approach to the task of classifying image tags on public available datasets.