Bayesian multinomial logistic-normal (MLN) models are popular for the analysis of sequence count data (e.g., microbiome or gene expression data) due to their ability to model multivariate count data with complex covar...
详细信息
Bayesian multinomial logistic-normal (MLN) models are popular for the analysis of sequence count data (e.g., microbiome or gene expression data) due to their ability to model multivariate count data with complex covariance structure. However, existing implementations of MLN models are limited to small datasets due to the non-conjugacy of the multinomial and logistic-normal distributions. Motivated by the need to develop efficient inference for Bayesian MLN models, we develop two key ideas. First, we develop the class of Marginally Latent Matrix-T Process (Marginally LTP) models. We demonstrate that many popular MLN models, including those with latent linear, non-linear, and dynamic linear structure are special cases of this class. Second, we develop an efficient inference scheme for Marginally LTP models with specific accelerations for the MLN subclass. Through application to MLN models, we demonstrate that our inference scheme are both highly accurate and often 4-5 orders of magnitude faster than MCMC.
This study explores the feasibility of deep learning for classifying nodule neoplasms, analyzing their performance on two openly available datasets, LUNGx SPIE, and LIDC-IDRI. These datasets offer valuable diversity i...
详细信息
1. Calibration is a crucial step for the validation of computational models and a challenging task to accomplish. 2. Dynamic Energy Budget (DEB) theory has experienced an exponential rise in the number of published pa...
详细信息
Presidential actions on Jan 20, 2025, by President Donald Trump, including executive orders, have delayed access to or led to the removal of crucial public health data sources in the USA. The continuous collection and...
详细信息
Multiple Instance Learning (MIL) is a widely employed framework for learning on gigapixel whole-slide images (WSIs) from WSI-level annotations. In most MIL-based analytical pipelines for WSI-level analysis, the WSIs a...
Multiple Instance Learning (MIL) is a widely employed framework for learning on gigapixel whole-slide images (WSIs) from WSI-level annotations. In most MIL-based analytical pipelines for WSI-level analysis, the WSIs are divided into patches, and deep features for patches (i.e., patch embeddings) are extracted prior to training to reduce the overall computational cost and cope with the GPUs’ limited RAM. Because of this bottleneck, incorporating patch-level data augmentations during training adds an extra computational burden. To overcome this limitation, we present EmbAugmenter, a data augmentation generative adversarial network (DA-GAN) that can synthesize data augmentations in the embedding space rather than in the pixel space, thereby significantly reducing the computational requirements. Experiments on the SICAPv2 dataset show that our approach outperforms MIL without augmentation and is on par with traditional patch-level augmentation for MIL training while being substantially faster.
The lowest time search in the dataset that E. Taillard utilized employs a heuristic approach based on tabu search techniques to get the predicted solution. Glover’s study gives a broad description of tabu search, whi...
The lowest time search in the dataset that E. Taillard utilized employs a heuristic approach based on tabu search techniques to get the predicted solution. Glover’s study gives a broad description of tabu search, which is commonly encountered in Taillard’s job shop scheduling difficulties and Widmer et al.’s flow shop sorting challenges. Although tabu search is relatively simple to use and typically yields excellent results, it takes a long time to complete. In this research a hybrid ACO and PSO was carried out to minimize makespan in the Job Shop Scheduling Problem which was used as sourced from benchmark data which is secondary data obtained from E. Taillard “Benchmarks for basic scheduling problems” which consists of job shop matrix data (job × machine) measuring 4 × 4, 5 × 5, 7 × 7, 10 × 10, 15 × 15, 20 × 20, 30 × 15, 30 × 20, 50 × 15 and 50 × 20. Hybrid is carried out by calculating the Pbest value, namely the process position of each job on the machine to get the best solution using the PSO algorithm. Next, calculate the Gbest (Global best) value for the position of each job on the best machine on the entire machine using the PSO algorithm and initialize the ACO parameters using the PBest and Gbest values. The results of research on datasets with sizes 10×10, 15×15, 20×20, 30×15, 30×20, 50×15 and 50×20 produce smaller makespan compared to the lower bound on the dataset with an average minimum makespan improvement value of 1.184.
With recent advances in artificial intelligence (AI) and robotics, unmanned vehicle swarms have received great attention from both academia and industry due to their potential to provide services that are difficult an...
详细信息
Accurate prediction of major histocompatibility complex (MHC)-peptide binding affinity can improve our understanding of cellular immune responses and guide personalized immunotherapies. Nevertheless, the existing deep...
详细信息
SARS CoV-2 is a fascinating topic to investigate, especially in Indonesia and Malaysia, which share similar racial demographics. However, statistical analysis of information on the SARS CoV-2 from a database, especial...
SARS CoV-2 is a fascinating topic to investigate, especially in Indonesia and Malaysia, which share similar racial demographics. However, statistical analysis of information on the SARS CoV-2 from a database, especially GISAID, does not contain specific customizations related to virus comparisons between selected countries. Therefore, the researchers conducted statistical analysis and data visualization using the Python programming language to describe and investigate SARS CoV-2 Indonesia and Malaysia from the GISAID database. SARS CoV-2 metadata from Indonesia (N=117) and Malaysia (N=250), which were gathered during 2020, were compared. This comparison was aimed to investigate the discrepancies of COVID-19 cases in closely related populations. Firstly, data visualization was conducted using the Python Matplotlib library to create bar charts for clades and mutation comparison. Additionally, a series of boxplots were generated to show age discrepancies stratified by gender. Furthermore, the statistical tests showed that only the dominant Malaysian (G and O) clades were found to be significantly different compared to Indonesian cases (p-value=0.016). The proportion of two major mutations (G614D and NSP12 P323L) were also significantly different in the two countries caused by the dominant clade differences (p-value=0.007). Lastly, the differences in the age distribution of COVID-19 cases between the two countries were significant only in the male group (p-value=0.017).
Support Vector Regression (SVR) is often used in forecasting. Adjustment of parameters in the SVR affects the results of forecasting. This study aims to analyze the SVR method that is optimized using Harris Hawks Opti...
详细信息
Support Vector Regression (SVR) is often used in forecasting. Adjustment of parameters in the SVR affects the results of forecasting. This study aims to analyze the SVR method that is optimized using Harris Hawks Optimization (HHO), hereinafter referred to as HHO-SVR. The HHO-SVR was evaluated using five benchmark datasets to determine the performance of this method. The HHO process is also compared based on the type of kernel and other metaheuristic algorithms. The results showed that the HHO-SVR has almost the same performance as other methods but is less efficient in terms of time. In addition, the type of kernel also affects the process and results.
暂无评论