In this paper, we describe our contributions to the challenge of detection and classification of acoustic scenes and events. We propose multi-scale convolutional recurrent neural network(Multi-scale CRNN), a novel weakly-supervised learning framework for sound event detection. By integrating information from different time resolutions, the multi-scale method can capture both the fine-grained and coarse-grained features of sound events and model the temporal dependency including fine-grained dependency and long-term dependency. Furthermore, the ensemble method proposed in the paper reduces the frame-level prediction errors using classification results. The proposed method achieves 29.2% in the event-based F1-score and 1.40 in event-based error rate in development set of DCASE2018 task4 compared to the baseline of 14.1% F-value and 1.54 error rate [1].
Importance Gastric and gastroesophageal junction cancers are diagnosed in more than 1 million people worldwide annually, and few effective treatments are available. Sintilimab, a recombinant human IgG4 monoclonal antibody that binds to programmed cell death 1 (PD-1), in combination with chemotherapy, has demonstrated promising efficacy. Objective To compare overall survival of patients with unresectable locally advanced or metastatic gastric or gastroesophageal junction cancers who were treated with sintilimab with chemotherapy vs placebo with chemotherapy. Also compared were a subset of patients with a PD ligand 1 (PD-L1) combined positive score (CPS) of 5 or more (range, 1-100). Design, Setting, and Participants Randomized, double-blind, placebo-controlled, phase 3 clinical trial conducted at 62 hospitals in China that enrolled 650 patients with unresectable locally advanced or metastatic gastric or gastroesophageal junction adenocarcinoma between January 3, 2019, and August 5, 2020. Final follow-up occurred on June 20, 2021. Interventions Patients were randomized 1:1 to either sintilimab (n = 327) or placebo (n = 323) combined with capecitabine and oxaliplatin (the XELOX regimen) every 3 weeks for a maximum of 6 cycles. Maintenance therapy with sintilimab or placebo plus capecitabine continued for up to 2 years. Main Outcomes and Measures The primary end point was overall survival time from randomization. Results Of the 650 patients (mean age, 59 years; 483 [74.3%] men), 327 were randomized to sintilimab plus chemotherapy and 323 to placebo plus chemotherapy. Among the randomized patients, 397 (61.1%) had tumors with a PD-L1 CPS of 5 or more; 563 (86.6%) discontinued study treatment and 388 (59.7%) died; 1 patient (<0.1%) was lost to follow-up. Among all randomized patients, sintilimab improved overall survival compared with placebo (median, 15.2 vs 12.3 months; stratified hazard ratio [HR], 0.77 [95% CI, 0.63-0.94]; P = .009). Among patients with a CPS of 5 or more, sintilimab improved overall survival compared with placebo (median, 18.4 vs 12.9 months; HR, 0.66 [95% CI, 0.50-0.86]; P = .002). The most common grade 3 or higher treatment-related adverse events were decreased platelet count (sintilimab, 24.7% vs placebo, 21.3%), decreased neutrophil count (sintilimab, 20.1% vs placebo, 18.8%), and anemia (sintilimab, 12.5% vs placebo, 8.8%). Conclusions and Relevance Among patients with unresectable locally advanced or metastatic gastric and gastroesophageal junction adenocarcinoma treated with first-line chemotherapy, sintilimab significantly improved overall survival for all patients and for patients with a CPS of 5 or more compared with placebo. Trial Registration ClinicalTrials.gov Identifier: NCT03745170
In this paper, we describe our ensemble-based system designed by guoym Team for the SemEval-2020 Task 8, Memotion Analysis. In our system, we utilize five types of representation of data as input of base classifiers to extract information from different aspects. We train five base classifiers for each type of representation using five-fold cross-validation. Then the outputs of these base classifiers are combined through data-based ensemble method and feature-based ensemble method to make full use of all data and representations from different aspects. Our method achieves the performance within the top 2 ranks in the final leaderboard of Memotion Analysis among 36 Teams.
Abstract Maize (Zea mays) kernel size is an important factor determining grain yield; although numerous genes regulate kernel development, the roles of RNA polymerases in this process are largely unclear. Here, we characterized the defective kernel 701 (dek701) mutant that displays delayed endosperm development but normal vegetative growth and flowering transition, compared to its wild type. We cloned Dek701, which encoded ZmRPABC5b, a common subunit to RNA polymerases I, II and III. Loss-of-function mutation of Dek701 impaired the function of all three RNA polymerases and altered the transcription of genes related to RNA biosynthesis, phytohormone response and starch accumulation. Consistent with this observation, loss-of-function mutation of Dek701 affected cell proliferation and phytohormone homeostasis in maize endosperm. Dek701 was transcriptionally regulated in the endosperm by the transcription factor Opaque2 through binding to the GCN4 motif within the Dek701 promoter, which was subjected to strong artificial selection during maize domestication. Further investigation revealed that DEK701 interacts with the other common RNA polymerase subunit ZmRPABC2. The results of this study provide substantial insight into the Opaque2–ZmRPABC5b transcriptional regulatory network as a central hub for regulating endosperm development in maize.
Abstract Epialleles, the heritable epigenetic variants that are not caused by changes in DNA sequences, can broaden genetic and phenotypic diversity and benefit to crop breeding, but very few epialleles related to agricultural traits have been identified in maize. Here, we cloned a small kernel mutant, smk-wl10, from maize, which encoded a tubulin-folding cofactor B (ZmTFCB) protein. Expression of the ZmTFCB gene decreased in the smk-wl10 mutant, which arrested embryo, endosperm and basal endosperm transfer layer developments. Overexpression of ZmTFCB could complement the defective phenotype of smk-wl10. No nucleotide sequence variation in ZmTFCB could be found between smk-wl10 and wild type (WT). Instead, we detected hypermethylation of nucleotide CHG (where H is A, C or T nucleotide) sequence contexts and increased level of histone H3K9me2 methylation in the upstream sequence of ZmTFCB in smk-wl10 compared with WT, which might respond to the attenuating transcription of ZmTFCB. In addition, yeast two-hybrid and bimolecular fluorescence complementation assays identified a strong interaction between ZmTFCB and its homolog ZmTFCE. Thus, our work identifies a novel epiallele of the maize ZmTFCB gene, which might represent a common phenomenon in the epigenetic regulation of important traits such as kernel development in maize.
The development of agriculture with plateau characteristics is a market-oriented strategic choice, made by Yunnan Province, of agricultural economy with regional characteristics, on the basis of resources and location advantages, as well as geographical division of the national economy. The characteristic agribusiness is an important carrier for building a new agricultural management system with plateau characteristics, and also a key way to promote characteristic agricultural industrialization. In this paper, with 26 agribusinesses with plateau characteristics in Yunnan Province as samples, we establish the competitiveness evaluation system for the agribusiness with plateau characteristics, and use the operating data (2012-2014) and AHP to calculate and the sample business competitiveness index and sort these businesses. Finally, we make a comprehensive analysis on the competitiveness of sample agribusinesses with plateau characteristics in Yunnan Province, in order to provide decision-making basis for promotion of the competitiveness of the agribusiness with plateau characteristics.
Online shoppers depend on customer reviews when evaluating products or services. However, in the international online marketplace, reviews in a user's language may not be available. Translation of online customer reviews is therefore an important service. A crucial aspect of this task is translating opinion words, key words that capture the reviewers' sentiments. This is challenging because opinion words often have multiple translations. We propose an unsupervised opinion word translation disambiguation scoring method using dependency distance and feature-opinion association as weighting factors. The scores of an opinion word's translation and its surrounding words' translations are estimated using Google snippets. We focus on Japanese-Chinese translation of hotel review s from Rakutan Travel, using the 10 most common polysemous Japanese opinion words to evaluate system performance. Results show our weighting factors significantly improve translation accuracy compared to Google and Excite.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)