72 research outputs found
Generative Adversarial Networks-Based AI/ML Model Adaptive Retraining for Beyond 5G Networks
No full textBeyond fifth-generation (B5G) networks aim to support high data rates, low-latency applications, and massive machine communications. Artificial Intelligence (AI) and Machine Learning (ML) can help to improve B5G network performance and efficiency. However, dynamic service demands of B5G cause AI/ML performance degradation, resulting in violations of Service Level Agreements (SLA), over- or under-provisioning of resources, etc. Retraining is essential to address the performance degradation of the AI/ML models. Existing threshold and periodic retraining approaches have potential disadvantages, such as SLA violations and inefficient resource utilization for setting a threshold parameter in a dynamic environment. This paper presents a novel algorithm that predicts when to retrain AI/ML models using the generative adversarial networks (GANs) architecture. The proposed predictive approach is evaluated for a Quality of Service (QoS) prediction use case on O-RAN Software Community (OSC) platform and compared to the predictive approach based on the classifier and the threshold approach. The results show that the proposed predictive approach outperforms both the classifier-based predictive and threshold approaches
IRS-1-Rad51 nuclear interaction sensitizes JCV T-antigen positive medulloblastoma cells to genotoxic treatment
No full textThe large T-antigen from human polyomavirus JC (JCV T-antigen) is suspected to play a role in malignant transformation. Previously, we reported that JCV T-antigen requires the presence of a functional insulin-like growth factor I receptor (IGF-IR) for transformation of fibroblasts and for survival of medulloblastoma cell lines; that IGF-IR is phosphorylated in medulloblastoma biopsies and that JCV T-antigen inhibits homologous recombination-directed DNA repair, causing accumulation of mutations. Here we are evaluating whether JCV T-antigen positive and negative mouse medulloblastoma cell lines, which significantly differ in their tumorigenic properties, are also different in their abilities to repair double strand breaks of DNA (DSBs). Our results show that despite much stronger tumorigenic potential, JCV T-antigen positive medulloblastoma cells are more sensitive to genotoxic agents (cisplatin and gamma-irradiation). Subsequent analysis of DNA repair of DSBs indicated that homologous recombination-directed DNA repair (HRR) was selectively attenuated in JCV T-antigen positive medulloblastoma cells. JCV T-antigen did not affect HRR directly. In the presence of JCV T-antigen, insulin receptor substrate 1 (IRS-1) translocated to the nucleus where it co-localized with Rad51, possibly attenuating HRR
Orchestrating edge- And cloud-based predictive analytics services
No full textIn the Zero-touch network and service management (ZSM) architecture, devised by ETSI, making predictions on the observed data is among the functions provided by the analytics block of the control loop cycle. Prediction performance depends on several parameters, such as the utilized computational resources, the leveraged prediction techniques, the deployment location of the prediction tools with respect to the data.This paper proposes a Hybrid Forecast Framework (HFF) running at the network edge or in the cloud to provide fore-casting with the performance required by the control loop cycle. Forecasting at the edge might shorten the control loop cycle if resources shall be made available locally where data is collected. However, in general, edge computational resources are less abundant than the cloud ones, thus causing longer time to perform the prediction. On the opposite, forecasting in the cloud might require more time for the data to reach the utilized tools but more computational resources could be exploited. The HFF is based on utilizing traditional time series analysis prediction algorithms to minimize the utilized resources and energy at the edge while it exploits AI/ML tools to make predictions in the cloud.Results show that for short lead time (i.e., the time, in the future, at which the status of the considered parameter is predicted) edge-based prediction exploiting time series analysis provides better accuracy, requires less resources and time (thus energy) than cloud-based prediction. However, if the lead time is long, cloud-based prediction exploiting Artificial Intelligence/Machine Learning (AI/ML) provides better accuracy. Thus, if the lead time is long, it is preferable because the long lead time compensates for the higher time for prediction, mainly due to data transfer
Functional correlates of positional and gender-specific renal asymmetry in drosophila
No full textAccordingly, the physical asymmetry of the tubules in the body cavity is directly adaptive. Now that the detailed machinery underlying internal asymmetry is starting to be delineated, our work invites the investigation, not just of tissues in isolation, but in the context of their unique physical locations and milieux
Recommended from our members
Diffusion bonding of large substrate MECS devices based on differential thermal expansion
No full textDiffusion bonding has been widely used within microlamination architectures for the
fabrication of Micro Energy and Chemical Systems (MECS). MECS are microsystems
with the ability to process bulk amounts of fluid within highly-parallel microchannel
arrays capable of accelerated heat and mass transfer. Thus far, diffusion bonding of
microchannel arrays is commonly done in a vacuum hot press system. The use of the hot
press greatly restricts the production rate due to vacuum pump-down time and heating-up
and cooling-down periods. Furthermore, larger substrates are gaining interest in the
system design of MECS devices and it is not apparent that uniaxial pressing within a
hydraulic vacuum hot press will provide the bonding pressure uniformity necessary for
large substrate bonding. This thesis presents a novel fabrication approach for the high-volume
thermal bonding of large substrate MECS devices with the use of controlled
thermal expansion. A large substrate thermal bonding clamp based on the principle of
differential thermal expansion was developed with focus on controlling the bonding
pressure magnitude, timing, sensitivity and uniformity
Mapping an atlas of tissue-specific drosophila melanogaster metabolomes by high resolution mass spectrometry
No full textMetabolomics can provide exciting insights into organismal function, but most work on simple models has focussed on the whole organism metabolome, so missing the contributions of individual tissues. Comprehensive metabolite profiles for ten tissues from adult Drosophila melanogaster were obtained here by two chromatographic methods, a hydrophilic interaction (HILIC) method for polar metabolites and a lipid profiling method also based on HILIC, in combination with an Orbitrap Exactive instrument. Two hundred and forty two polar metabolites were putatively identified in the various tissues, and 251 lipids were observed in positive ion mode and 61 in negative ion mode. Although many metabolites were detected in all tissues, every tissue showed characteristically abundant metabolites which could be rationalised against specific tissue functions. For example, the cuticle contained high levels of glutathione, reflecting a role in oxidative defence; the alimentary canal (like vertebrate gut) had high levels of acylcarnitines for fatty acid metabolism, and the head contained high levels of ether lipids. The male accessory gland uniquely contained decarboxylated S-adenosylmethionine. These data thus both provide valuable insights into tissue function, and a reference baseline, compatible with the FlyAtlas.org transcriptomic resource, for further metabolomic analysis of this important model organism, for example in the modelling of human inborn errors of metabolism, aging or metabolic imbalances such as diabetes
Interesting plant records from Visakhapatnam District, Andhra Pradesh, India
No full textAn extensive botanical survey conducted in Paderu, Narsipatnam and Chintapalli forest areas in Visakhapatnam District of Andhra Pradesh, resulted in a record of 18 interesting angiosperm species including two distributional records for south India (Cynanchum corymbosum and Raphistemma pulchellum); two (Cassine paniculata and Cleidion javanicum) for Andhra Pradesh as well as Eastern Ghats; seven additions to Visakhapatnam district and seven others are endemic and rare species. All species are presented here with brief botanical descriptions and notes
Evaluation of ginger (Zingiber officinale Rosc.) varieties in high altitude and tribal zone of Srikakulam district of Andhra Pradesh
No full textThe yield performance and simple association between yield and its components were studied in eight varieties of ginger during kharif 2007 and 2008. The variety Suprabha was taller and recorded more number of leaves, tillers plant-1 and number of finger rhizomes plant-1. It produced significantly more fresh rhizome yield of 21.71 t ha-1 than all the other the varieties tested. Among the varieties Chintapalli local produced more number of mother rhizomes plant-1. Number of tillers plant-1, number of mother and finger rhizomes plant-1 and fresh rhizome yield showed high GCV, PCV, heritability and genetic advance as per cent mean. The simple correlation studies indicated that number of tillers plant-1, number of mother and finger rhizomes plant-1 recorded highly significant association with yield.
 
Embedding theorems on hyperelliptic varieties
No full textIn this paper, we investigate linear systems on hyperelliptic varieties. We prove analogues of well-known theorems on abelian varieties, like Lefschetz’s embedding theorem and higher k-jet embedding theorems. Syzygy or Np-properties are also deduced for appropriate powers of ample line bundles. This is a first result on linear series, on hyperelliptic varieties
Evaluation of ginger (Zingiber officinale Rosc.) varieties in high altitude and tribal zone of Srikakulam district of Andhra Pradesh
No full textThe yield performance and simple association between yield and its components were studied in eight varieties of ginger during kharif 2007 and 2008. The variety Suprabha was taller and recorded more number of leaves, tillers plant-1 and number of finger rhizomes plant-1. It produced significantly more fresh rhizome yield of 21.71 t ha-1 than all the other the varieties tested. Among the varieties Chintapalli local produced more number of mother rhizomes plant-1. Number of tillers plant-1, number of mother and finger rhizomes plant-1 and fresh rhizome yield showed high GCV, PCV, heritability and genetic advance as per cent mean. The simple correlation studies indicated that number of tillers plant-1, number of mother and finger rhizomes plant-1 recorded highly significant association with yield.
 
- …
