similar domain
Recently Published Documents


TOTAL DOCUMENTS

40
(FIVE YEARS 14)

H-INDEX

13
(FIVE YEARS 2)

Mathematics ◽  
2021 ◽  
Vol 10 (1) ◽  
pp. 112
Author(s):  
Muhammad S. Battikh ◽  
Artem A. Lenskiy

Reconstruction-based approaches to anomaly detection tend to fall short when applied to complex datasets with target classes that possess high inter-class variance. Similar to the idea of self-taught learning used in transfer learning, many domains are rich with similar unlabeled datasets that could be leveraged as a proxy for out-of-distribution samples. In this paper we introduce the latent-insensitive autoencoder (LIS-AE) where unlabeled data from a similar domain are utilized as negative examples to shape the latent layer (bottleneck) of a regular autoencoder such that it is only capable of reconstructing one task. We provide theoretical justification for the proposed training process and loss functions along with an extensive ablation study highlighting important aspects of our model. We test our model in multiple anomaly detection settings presenting quantitative and qualitative analysis showcasing the significant performance improvement of our model for anomaly detection tasks.


2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Subhendu Paul ◽  
Emmanuel Lorin

AbstractWe derive a novel model escorted by large scale compartments, based on a set of coupled delay differential equations with extensive delays, in order to estimate the incubation, recovery and decease periods of COVID-19, and more generally any infectious disease. This is possible thanks to some optimization algorithms applied to publicly available database of confirmed corona cases, recovered cases and death toll. In this purpose, we separate (1) the total cases into 14 groups corresponding to 14 incubation periods, (2) the recovered cases into 406 groups corresponding to a combination of incubation and recovery periods, and (3) the death toll into 406 groups corresponding to a combination of incubation and decease periods. In this paper, we focus on recovery and decease periods and their correlation with the incubation period. The estimated mean recovery period we obtain is 22.14 days (95% Confidence Interval (CI) 22.00–22.27), and the 90th percentile is 28.91 days (95% CI 28.71–29.13), which is in agreement with statistical supported studies. The bimodal gamma distribution reveals that there are two groups of recovered individuals with a short recovery period, mean 21.02 days (95% CI 20.92–21.12), and a long recovery period, mean 38.88 days (95% CI 38.61–39.15). Our study shows that the characteristic of the decease period and the recovery period are alike. From the bivariate analysis, we observe a high probability domain for recovered individuals with respect to incubation and recovery periods. A similar domain is obtained for deaths analyzing bivariate distribution of incubation and decease periods.


2021 ◽  
Vol 8 ◽  
Author(s):  
Yutao Huang ◽  
Zijian Jiang ◽  
Xiangyu Gao ◽  
Peng Luo ◽  
Xiaofan Jiang

Armadillo repeat-containing proteins (ARMCs) are widely distributed in eukaryotes and have important influences on cell adhesion, signal transduction, mitochondrial function regulation, tumorigenesis, and other processes. These proteins share a similar domain consisting of tandem repeats approximately 42 amino acids in length, and this domain constitutes a substantial platform for the binding between ARMCs and other proteins. An ARMC subfamily, including ARMC1∼10, ARMC12, and ARMCX1∼6, has received increasing attention. These proteins may have many terminal regions and play a critical role in various diseases. On the one hand, based on their similar central domain of tandem repeats, this ARMC subfamily may function similarly to other ARMCs. On the other hand, the unique domains on their terminals may cause these proteins to have different functions. Here, we focus on the ARMC subfamily (ARMC1∼10, ARMC12, and ARMCX1∼6), which is relatively conserved in vertebrates and highly conserved in mammals, particularly primates. We review the structures, biological functions, evolutions, interactions, and related diseases of the ARMC subfamily, which involve more than 30 diseases and 40 bypasses, including interactions and relationships between more than 100 proteins and signaling molecules. We look forward to obtaining a clearer understanding of the ARMC subfamily to facilitate further in-depth research and treatment of related diseases.


Cells ◽  
2021 ◽  
Vol 10 (9) ◽  
pp. 2331
Author(s):  
Igor Yu. Dolmatov ◽  
Vladimir A. Nizhnichenko ◽  
Lyudmila S. Dolmatova

Echinoderms are one of the most ancient groups of invertebrates. The study of their genomes has made it possible to conclude that these animals have a wide variety of matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs). The phylogenetic analysis shows that the MMPs and TIMPs underwent repeated duplication and active divergence after the separation of Ambulacraria (Echinodermata+Hemichordata) from the Chordata. In this regard the homology of the proteinases and their inhibitors between these groups of animals cannot be established. However, the MMPs of echinoderms and vertebrates have a similar domain structure. Echinoderm proteinases can be structurally divided into three groups—archetypal MMPs, matrilysins, and furin-activatable MMPs. Gelatinases homologous to those of vertebrates were not found in genomes of studied species and are probably absent in echinoderms. The MMPs of echinoderms possess lytic activity toward collagen type I and gelatin and play an important role in the mechanisms of development, asexual reproduction and regeneration. Echinoderms have a large number of genes encoding TIMPs and TIMP-like proteins. TIMPs of these animals, with a few exceptions, have a structure typical for this class of proteins. They contain an NTR domain and 10–12 conservatively located cysteine residues. Repeated duplication and divergence of TIMP genes of echinoderms was probably associated with an increase in the functional importance of the proteins encoded by them in the physiology of the animals.


2021 ◽  
Vol 9 (3) ◽  
pp. 43
Author(s):  
Gidon T. Frischkorn ◽  
Claudia C. von Bastian

Process-Overlap Theory (POT) suggests that measures of cognitive abilities sample from sets of independent cognitive processes. These cognitive processes can be separated into domain-general executive processes, sampled by the majority of cognitive ability measures, and domain-specific processes, sampled only by measures within a certain domain. According to POT, fluid intelligence measures are related because different tests sample similar domain-general executive cognitive processes to some extent. Re-analyzing data from a study by De Simoni and von Bastian (2018), we assessed domain-general variance from executive processing tasks measuring inhibition, shifting, and efficiency of removal from working memory, as well as examined their relation to a domain-general factor extracted from fluid intelligence measures. The results showed that domain-general factors reflecting general processing speed were moderately and negatively correlated with the domain-general fluid intelligence factor (r = −.17–−.36). However, domain-general factors isolating variance specific to inhibition, shifting, and removal showed only small and inconsistent correlations with the domain-general fluid intelligence factor (r = .02–−.22). These findings suggest that (1) executive processing tasks sample only few domain-general executive processes also sampled by fluid intelligence measures, as well as (2) that domain-general speed of processing contributes more strongly to individual differences in fluid intelligence than do domain-general executive processes.


Sensor Review ◽  
2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Sudarsana Jena ◽  
Ankur Gupta

Purpose Considering its vast utility in industries, this paper aims to present a detailed review on fundamentals, classification and progresses in pressure sensors, along with its wide area of applications, its design aspects and challenges, to provide state-of-the-art gist to the researchers of the similar domain at one place. Design/methodology/approach Swiftly emerging research prospects in the micro-electro-mechanical system (MEMS) enable to build complex and sophisticated micro-structures on a substrate containing moving masses, cantilevers, flexures, levers, linkages, dampers, gears, detectors, actuators and many more on a single chip. One of the MEMS initial products that emerged into the micro-system technology is MEMS pressure sensor. Because of their high performance, low cost and compact in size, these sensors are extensively being adopted in numerous applications, namely, aerospace, automobile and bio-medical domain, etc. These application requirements drive and impose tremendous conditions on sensor design to overcome the tedious design and fabrication procedure before its reality. MEMS-based pressure sensors enable a wide range of pressure measurement as per the application requirements. Findings The paper provides a detailed review on fundamentals, classification and progresses in pressure sensors, along with its wide area of applications, its design aspects and challenges, to provide state of the art gist to the researchers of the similar domain at one place. Originality/value The present paper discusses the basics of MEMS pressure sensors, their working principles, different design aspects, classification, type of sensing diaphragm used and illustration of various transduction mechanisms. Moreover, this paper presents a comprehensive review on present trend of research on MEMS-based pressure sensors, its applications and the research gap observed till date along with the scope for future work, which has not been discussed in earlier reviews.


2020 ◽  
Author(s):  
Patricia Balthazar ◽  
Scott Jeffery Lee ◽  
Daniel Rubin ◽  
Terry Dessar ◽  
Judy Gichoya ◽  
...  

Transfer learning is a common practice in image classification with deep learning where the available data is often limited for training a complex model with millions of parameters. However, transferring language models requires special attention since cross-domain vocabularies (e.g. between news articles and radiology reports) do not always overlap as the pixel intensity range overlaps mostly for images. We present a concept of similar domain adaptation where we transfer an interinstitutional language model between two different modalities (ultrasound to MRI) to capture liver abnormalities. Our experiments show that such transfer is more effective for performing shared targeted task than generic language space transfer. We use MRI screening exam reports for hepatocellular carcinoma as the use-case and apply the transfer language space strategy to automatically label thousands of imaging exams.


Author(s):  
Farzaneh Shoeleh ◽  
Mohammad Mehdi Yadollahi ◽  
Masoud Asadpour

Abstract There is an implicit assumption in machine learning techniques that each new task has no relation to the tasks previously learned. Therefore, tasks are often addressed independently. However, in some domains, particularly reinforcement learning (RL), this assumption is often incorrect because tasks in the same or similar domain tend to be related. In other words, even though tasks are quite different in their specifics, they may have general similarities, such as shared skills, making them related. In this paper, a novel domain adaptation-based method using adversarial networks is proposed to do transfer learning in RL problems. Our proposed method incorporates skills previously learned from source task to speed up learning on a new target task by providing generalization not only within a task but also across different, but related tasks. The experimental results indicate the effectiveness of our method in dealing with RL problems.


Within the domain of image processing the applications of the fractal image compression through the affine transform is rarely implemented. The major drawback it consists is its time consumptions in simulations as well as in hardware implementation. The reason behind this is its comparatively matching the range blocks with the varied number of domain blocks. Additionally, the compression ration obtained from this sets of techniques are much more that the other similar working techniques. In order to overcome, this drawback, here a concept of intensity mismatch searching techniques is intruded within this particular paper. While in the mean time, it only concluded in much reduce in time of searching for similar domain block form a pool. And also this technique does not affect within the compression ratio and PSNR measurement form the retrieved image. Further benefitted in hardware applications where the matters for time consumptions is a major factor.


2019 ◽  
Vol 8 (3) ◽  
pp. 3469-3473

Employee engagement is the need for modern business; it goes beyond employee satisfaction. It is an emotional commitment by employees towards their company. It has a direct influence on efficiency, productivity, employee retention, and the company’s profitability. Employee surveys are the most commonly used method to gather wide-ranging information from the employees about how satisfied and how engaged they are. Some of them also support in defining action plans to increase employee engagement results. However, it is a reactive measure to define and act on corrective actions, based employee satisfaction survey. In this research, I would like to study all secondary data available/accessible for different business domains in India and develop the guiding principles for new entrepreneurs. This will consist of tried, tested, and successful best practices from other companies of a similar domain.


Sign in / Sign up

Export Citation Format

Share Document