Change point detection methods try to find any sudden changes in the patterns and features of a given time series. In this paper a new change point detection method is presented, where the window width is automaticall...
详细信息
Applying evidence-based medicine prevents medical errors highlighting the need for applying Clinical Guidelines (CGs) to improve patient care by nurses. However, nurses often face challenges in utilizing CGs due to pa...
详细信息
We present AircraftVerse, a publicly available aerial vehicle design dataset. Aircraft design encompasses different physics domains and, hence, multiple modalities of representation. The evaluation of these cyber-phys...
详细信息
software testing is very important in software development to ensure its quality and reliability. As softwaresystems have become more complex, the number of test cases has increased, which presents the challenge of e...
详细信息
This comprehensive review starts with diving into the progress and real-world applications of combining multi-omics data analysis with machine learning techniques in cancer research. Multi-omics involves examining var...
This comprehensive review starts with diving into the progress and real-world applications of combining multi-omics data analysis with machine learning techniques in cancer research. Multi-omics involves examining various biological data types like genomics, transcriptomics, proteomics, and metabolomics together to enhance our understanding of complex biological systems. By merging machine learning with multiomics data, we highlight the advantages for cancer studies, the deeper insights they yield and increased performance and results. Furthermore, we explore existing literature that showcases the integration of multi-omics and machine learning in cancer research. As part of our study, we conduct an experiment utilizing a multiomics dataset to predict the survival of breast cancer patients. We compare three distinct machine learning methods-ensemble, DeepProg, and DCAP-for survival prediction and conclude that despite the ensemble method that increased the prediction results of DeepProg over DCAP in multi-model setting, but the primitive capacity for DCAP is better in single model setting and achieves higher accuracy than DeepProg with noticeable margin 0.628 to 0.57 on C-Index metric, which strongly recommends using Denoising Autoencoder as the base for dimensionality reduction over the vanilla Autoencoder. Another empirical results conclude that using gaussian mixture model with diagonal covariance matrix for Clustering, which is used in DeepProg, might hinder the process for identifying reasonable clusters due to the assumption of no or zero correlation between different features which might not hold true in our problem.
This paper presents a novel hybrid model comprising Evolutionary Scale Modeling (ESM), Convolution Neural Network (CNN) and Long Short Term Memory (LSTM) network for prediction of protein secondary structures from coi...
详细信息
ISBN:
(数字)9798331519094
ISBN:
(纸本)9798331519100
This paper presents a novel hybrid model comprising Evolutionary Scale Modeling (ESM), Convolution Neural Network (CNN) and Long Short Term Memory (LSTM) network for prediction of protein secondary structures from coil (C), helix (H), and sheet (E)—from amino acid sequences using deep learning techniques. Each architecture leverages unique strengths, with LSTMs capturing long-range dependencies, CNNs extracting local spatial patterns, and ESM enhancing contextual understanding of sequences. The hybrid model was trained and tested using two key datasets: the UniProt dataset and the pdb-intersect-pisces dataset, which provide a rich source of protein sequences and structural information. The proposed model achieved an accuracy of 89.22%, demonstrating robust performance in protein secondary structure prediction.
Ceylon tea industry faces a major challenge in the form of pathogen-induced crop loss, with Blister Blight (BB) caused by Exobasidium vexans posing the greatest threat, leading to harvest losses of over 30%. This fung...
详细信息
Multi-pushdown systems are a standard model for concurrent recursive programs, but they have an undecidable reachability problem. Therefore, there have been several proposals to underapproximate their sets of runs so ...
详细信息
Human Posture Detection (HPD) and classification are the initial steps toward various computer vision applications related to security, advertisement, and healthcare. Even though more than three decades have passed, m...
详细信息
Human Posture Detection (HPD) and classification are the initial steps toward various computer vision applications related to security, advertisement, and healthcare. Even though more than three decades have passed, most methods have been focused only on detecting standing people. However, in real applications, human postures may be significantly different, such as standing, sitting, lying, and crouching, and the shapes of a human are varied with viewpoints, thus making detection and classification difficult. Moreover, the gradual transition from one posture to another further complicates the determination of the number of postures to be classified. This paper uses a pre-trained model to determine the number of unique human postures in activities. We approached PC as a multiclass detection problem using the very deep VGG 16 Convolutional Neural Network (CNN) empowered by an attention mechanism. Extract frames from video data recorded, followed by the derivation of features through a convolutional neural network. Classify different postures—bending, exercise, lying, sitting, and standing—with VGG 16 and attention mechanism. It is trained on the ETRI-Activity3D-LivingLab Dataset. Our proposed (VEACT-CNN) framework has shown stability in HPD with an overall accuracy of 95% and an F-measure of 98% while generating a very low level of false alarms rate (FAR) of 0.02 (2%). For comparison purposes, the accuracy obtained by other models on the same dataset is VGG 16.
This workshop is designed to intentionally share experiences and connect participants to their greatest asset, their voice. There is power in your voice. Engaging interaction, thought-provoking storytelling, and authe...
ISBN:
(数字)9798350372915
ISBN:
(纸本)9798350372922
This workshop is designed to intentionally share experiences and connect participants to their greatest asset, their voice. There is power in your voice. Engaging interaction, thought-provoking storytelling, and authentic truth telling will open the way for understanding the value of harnessing voice power when navigating professionally and personally. Despite diversity, equity, and inclusion efforts, stated commitments to diversity, the demographic characteristics of the professoriate and industry look remarkably like the faculty and professionals of 50 years ago for Blacks in computing. Blacks in computing constantly face discrimination in the form of microaggressions and white fragility. This discrimination has not translated to greater Black faculty and professional representation in their career advancement. Instead, the health of both faculty and professionals are being adversely impacted, as they are experiencing insomnia, anxiety, depression, emotional distress, and physical exhaustion. This disruption of norms has a negatively impacted the productivity of both Black men and women at work and can lead to their demotion and dismissal. Many times, these struggles are not shared, and their voices are silenced. This workshop primarily addresses how Blacks in computing can use their voice to advance their careers and utilize proven voice positioning strategies to harness their voice power and to maximize professional and personal wellbeing. These voice strategies will facilitate both Black men and women in their career advancement, skill development, and preparation for entry into their next career level. A failure to address these challenges has negative implications for both academia and industry
暂无评论