Statistics and Data Analysis in Geology

A Survey on Incorporating Domain Knowledge into Deep Learning for Medical Image Analysis

Parametric deep energy approach for elasticity accounting for strain gradient effects

Convolutional Neural Networks (CNN) have gained great success in many artificial intelligence tasks However, finding a good set of hyperparameters for a CNN remains a challenging task It usually takes an expert with deep knowledge, and trials and errors Genetic algorithms have been used in hyperparameter optimizations However, traditional genetic algorithms with fixed-length chromosomes may not be a good fit for optimizing deep learning hyperparameters, because deep learning models have variable number of hyperparameters depending on the model depth As the depth increases, the number of hyperparameters grows exponentially, and searching becomes exponentially harder It is important to have an efficient algorithm that can find a good model in reasonable time In this article, we propose to use a variable length genetic algorithm (GA) to systematically and automatically tune the hyperparameters of a CNN to improve its performance Experimental results show that our algorithm can find good CNN hyperparameters efficiently It is clear from our experiments that if more time is spent on optimizing the hyperparameters, better results could be achieved Theoretically, if we had unlimited time and CPU power, we could find the optimized hyperparameters and achieve the best results in the future

Efficient Hyperparameter Optimization in Deep Learning Using a Variable Length Genetic Algorithm.

In recent years, machine learning and deep learning have become popular methods for financial data analysis, including financial textual data, numerical data, and graphical data. This paper proposes to use sentiment analysis to extract useful information from multiple textual data sources and a blending ensemble deep learning model to predict future stock movement. The blending ensemble model contains two levels. The first level contains two Recurrent Neural Networks (RNNs), one Long-Short Term Memory network (LSTM) and one Gated Recurrent Units network (GRU), followed by a fully connected neural network as the second level model. The RNNs, LSTM, and GRU models can effectively capture the time-series events in the input data, and the fully connected neural network is used to ensemble several individual prediction results to further improve the prediction accuracy. The purpose of this work is to explain our design philosophy and show that ensemble deep learning technologies can truly predict future stock price trends more effectively and can better assist investors in making the right investment decision than other traditional methods.

A Novel Ensemble Deep Learning Model for Stock Prediction Based on Stock Prices and News

Adaptive computation offloading and resource allocation strategy in a mobile edge computing environment

Different biomedical computing methods for cancer-specific gene recognition have been developed in recent years. Currently, building an open-box machine learning system to discover explainable knowledge from gene expression data is a difficult research problem due to a large number of genes, a small number of samples, and noise. Fuzzy systems can be used to deal with data ambiguity and noise issues and extract meaningful knowledge from gene data. In this article, we create a new deep fuzzy neural network to handle the uncertainty in gene data to generate useful knowledge for specific disease diagnosis. A new hybrid algorithm is designed to preprocess data and select informative genes for accurate cancer detection. Various experiments using six different cancer datasets indicate that the new method has better and more reliable performance than the other conventional classification methods with different gene selection methods.

Deep Fuzzy Neural Networks for Biomarker Selection for Accurate Cancer Detection

As the complexity of deep learning models grows, the difficulty and time of training them also increases. Depending on the task, the complexity of the model, and the hardware resources available, the amount of training time could take hours, weeks or even months. To decrease the training time, we propose a method to intelligently freeze layers during the training process. Our method involves designing a formula to calculate normalized gradient differences for all layers with weights in the model, and then use the calculated values to decide how many layers should be frozen. We implemented our method on top of stochastic gradient descent, and performed experiments on standard image classification dataset CIFAR-10. Results show that our method can accelerate training on VGG nets, ResNets, and DenseNets while having similar test accuracy.

Fast Deep Learning Training through Intelligently Freezing Layers

Asphyxia is a respiratory injury that leads to a serious damage for infants. Early detection of asphyxia using Artificially Intelligent technology helps in reducing infant mortality rate when compared to traditional medical diagnosis, which is time consuming. In this paper, we propose a novel method through generating weighted prosodic features combined with acoustic features to form a merged feature matrix to classify asphyxiated baby crying effectively. The weights of the prosodic features are trained at the frame level with labeled data and can be optimized using deep learning approach with neural networks. The novel merged feature matrix is established with both acoustic and weighted prosodic features. The matrix has good ability to capture the diversity of variations within infant cries, especially for asphyxiated samples. Our method has the benefits of keeping the robustness and resolution of the classification model simultaneously. The effectiveness of this approach is evaluated on Baby Chillanto Database. Our method yields a significant reduction of 3.11%, 3.23%, and 1.43% absolute classification error rate compared with the results using single acoustic features, single prosodic features, and both acoustic and prosodic features, respectively. The testing accuracy in our method reaches 96.74%, which outperforms all other related studies on asphyxiated baby crying classification.

Xueli Xiao

Papers

Adaptive computation offloading and resource allocation strategy in a mobile edge computing environment

Efficient Hyperparameter Optimization in Deep Learning Using a Variable Length Genetic Algorithm.

Deep Fuzzy Neural Networks for Biomarker Selection for Accurate Cancer Detection

Fast Deep Learning Training through Intelligently Freezing Layers

Deep Learning for Asphyxiated Infant Cry Classification Based on Acoustic Features and Weighted Prosodic Features