60 research outputs found

    Neural network architectures for PP attachment disambiguation

    No full text
    Thesis: S.M. in Computer Science and Engineering, Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2014.This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.25Cataloged from student-submitted PDF version of thesis.Includes bibliographical references (pages 41-44).This thesis addresses the problem of Prepositional Phrase (PP) attachment disambiguation, a key challenge in syntactic parsing. In natural language sentences, a PP may often be attached to several possible candidates. While humans can usually identify the correct candidate successfully, syntactic parsers are known to have high error rated on this kind of construction. This work explores the use of compositional models of meaning in choosing the correct attachment location. The compositional model is defined using a recursive neural network. Word vector representations are obtained from large amounts of raw text and fed into the neural network. The vectors are first forward propagated up the network in order to create a composite representation, which is used to score all possible candidates. In training, errors are back-propagated down the network such that the composition matrix is updated from the supervised data. Several possible neural architectures are designed and experimentally tested in both English and Arabic data sets. As a comparative system, we offer a learning-to-rank algorithm based on an SVM classifier which has access to a wide range of features. The performance of this system is compared to the compositional models.by Yonatan Belinkov.S.M. in Computer Science and Engineerin

    Erratum: What is the source? Identifying the cause of septic shock in a patient (Annals of the American Thoracic Society (2020) 17 (236-239) DOI: 10.1513/AnnalsATS.201907-562CC)

    No full text
    Copyright © 2020 by the American Thoracic Society. The authors would like to correct the author order in their article published in the February 2020 issue of AnnalsATS (1). The correct author order should be as follows: Nimrita Sidhu, Noor Ul-Ain Baloch, and Yonatan Greenstein. For the convenience of our readers, AnnalsATS is replacing the online version of the article with a corrected version; we apologize for the error

    Penentuan Distribusi Magnet Sekitar Koil Melingkar Berarus: Suatu Pendekatan Numerik

    No full text
    Telah dilakukan penelitian untuk merancang dan membuat suatu aplikasi untuk penentuan distribusi magnet sekitar koil melingkar berarus melalui pendekatan numerik. Metode penelitian adalah membangun suatu model matematis pencarian nilai medan magnet pada titik titik di sekitar  suatu koil berarus. Penentuannya menggunakan hukum hukum elektromagnetika. Solusi matetmatis persamaan persamaan tersebut diperoleh dengan menggunakan pendekatan metode numerik. Model pendekatan numerik tersebut tersebut kemudian diimplementasikan dengan membuat suatu aplikasi komputer menggunakan bahasa C++. Selanjutnya dilakukan pengujian dengan menggunakan model analitik dan hasil hasil kualitatif eksperimental. Hasil penelitian menujukkan model yang digunakan dan implementasinya dengan program komputer memberikan hasil yang mirip dengan teori dan hasil kualitatif eksperimen. Kata kunci: koil berarus; distribusi magnet; bahasa C++; metode numeri

    LoRaWAN Class B Multicast Scalability

    No full text
    LoRaWAN has emerged as a popular IoT commu- nications technology. It comes with three classes of operation: A, B, and C. Although many IoT use-cases, like Firmware-over- the-Air updates, require multicast, Class A cannot be used for that purpose. Class C can, but consumes a lot of energy. This leaves Class B. In this paper, we investigate Class B multicast and its scalability properties. Issues like multicast member capacity, beacon blocking, and beacon collisions are highlighted, and several approaches to mitigate them are proposed: (1) “Ping-Slot Relaying,” to allow for more multicast members, (2) a scheduling approach indicating when to best send multicast packets, and (3) “Dynamic Region Formation” to coordinate the sending of beacons over multiple gateways. The proposed solutions do not require any modifications to the LoRaWAN protocol.Virtual/online event due to COVID-19 accepted author manuscriptEmbedded System

    Performance Analysis of Light Arrester (LA) Protecting Power Transformer at Distribution Substation MH284 Buha Manado

    No full text
    North Sulawesi (Manado) is located in the equatorial region with a tropical climate and high humidity, causing the density of lightning strikes in this area to be quite large yearly.  Lightning strikes can cause overvoltage, which can endanger the insulator on the line as well as other electrical equipment if allowed to flow into the system and channel to the load. Therefore, a protection system is needed to handle these disturbances. One of them is by installing a Light Arrester (LA).  Light Arrester is safety equipment from overvoltage disturbances by surges lightning or surges switching . The arrester works at a specific voltage above the operating voltage to dispose of the electric charge from surges lightning or surges switching. Under normal conditions, the light arrester acts as insulation. However, if there is any available surges lightning or surges switching , the arrester acts as a conductor that functions to pass high currents to the ground. The failure rate of arrester protection is highly dependent on the Basic Insulation Level (TID) of the equipment, the working voltage of the lightning arrester, and the location of the light arrester placement itself

    Model Sistem Informasi Biaya Kunjungan Wisata di Provinsi Sulawesi Utara

    No full text
    Cost is one of the important information needed by tourists and prospective tourists in planning and conducting tourist visits. This study aims to produce a system model that can record, process, and present information about the cost picture when tourists or potential tourists to travel in North Sulawesi province such as overview of accommodation costs, transportation, culinary and cost of tourist attraction. Using waterfall method in system development. Data collection by observation, interview and literature. Testing system using blackbox method. The result of this research is a system model that can present information about the cost picture starting from the description of minimum cost and others which make the information need of the cost can be fulfilled, so it is expected to help the tourist in traveling and support the tourism industry in North Sulawesi Province

    Pentecostalism and Alternative Paths for Self-Accomplishment in Kenya

    No full text
    Author : Daryona. This file is licensed under the Creative Commons Attribution Share Alike Introduction Yvan Droz, Senior Lecturer in Anthropology and Sociology at the Graduate Institute, and Yonatan N. Gez, PhD alumnus of the Institute and a member of the Martin Buber Society of Fellows in the Humanities and Social Sciences at HUJI Jerusalem, are among the authors of a chapter on Pentecostalism and self-accomplishment in Kenya, published in Religion and Human Security in Africa (Routl..

    LoRaWAN Class B Multicast: Scalability

    No full text
    LoRaWAN is amongst the most prominent LPWAN (Low Power Wide Area Networking) technologies and provides an operation mode, called Class B, where devices can be reached with limited latency. One of the most important features of LoRaWAN Class B is multicast, which allows a single transmission to be shared by multiple multicast group member end-devices. This enables key IoT functionalities such as FOTA (Firmware Over The Air) and other applications that need to transfer multiple packets to a setof end-devices. This thesis will delve into LoRaWAN Class B multicast and study its scalability -- the most important behavior to study in all IoT technologies. An ns-3 module for LoRaWAN Class B multicast has been extended and used for in-depth analysis and drawing scalability insights. The study also explores methods to improve performance as well as enhance scalability.Electrical Engineering | Embedded System
    corecore