444 research outputs found
sj-jpg-1-gsj-10.1177_21925682211041968 – Supplemental Material for Predictors of Prolonged Opioid Use After Lumbar Fusion and the Effects of Opioid Use on Patient-Reported Outcome Measures
No full textSupplemental Material, sj-jpg-1-gsj-10.1177_21925682211041968 for Predictors of Prolonged Opioid Use After Lumbar Fusion and the Effects of Opioid Use on Patient-Reported Outcome Measures by Jose A Canseco, Michael Chang, Brian A Karamian, Jennifer Z Mao, Ariana A ReyesMD, John Mangan, Srikanth N Divi, Dhruv KC Goyal, Harold I Salmons, Nicolas Dohse, Noah Levy, Maxwell Detweiler, D Greg Anderson, Jeffrey A Rihn, Mark F Kurd, Alan S Hilibrand, Christopher K Kepler, Alexander R Vaccaro and Gregory D Schroeder in Global Spine Journal</p
Dynamics of Network Formation Processes in the Co-Author Model
Get PDFThis article studies the dynamics in the formation processes of a mutual consent network in game theory setting: the Co-Author Model. In this article, a limited observation is applied and analytical results are derived. Then, 2 parameters are varied: the number of individuals in the network and the initial probability of the links in the network in its initial state. A simulation result shows a finding that is consistent with an analytical result for a state of equilibrium while it also shows different possible equilibria.Dynamics, Network, Game Theory, Model,Simulation, Equilibrium, Complexity
Assessing the Efficacy of Synthetic Data for Enhancing Machine Translation Models in Low Resource Domains
No full textAn artificially generated dataset mimics real-world data in terms of its statistical properties, but it contains no real information. Data around rare occurrences like Covid-19 pandemic is difficult to capture in real-world data due to their infrequent nature. Additionally, cost involved and time-consumption to gather real world data is a big challenge. In such cases, synthetic data can help create more balanced datasets for model training. This project investigates the effectiveness of using synthetic data for tuning machine translation models when training data is limited. The Covid-19 domain is chosen considering the urgency and importance of the global accessibility of information related to the pandemic. TICO-19, a publically available dataset was effectively formulated to cater to this need. The medical terminologies were extracted and passed to OpenAI API to generate training language pair data. The fine-tuned davinci model is then verified with blind test data provided under TICO-19 for translation from English to French. SacreBLEU score is used to compute the translation quality, the fine-tuned model has a significantly higher BLEU score of 19.54 in comparison to the base model with a BLEU score of 0.44. The adapted model also has a comparable score to the next-generation version of davinci with a BLEU score of 22.29
Isolated propeller aeroacoustics at positive and negative thrust Author links open overlay panel
Get PDFUsing propellers in negative thrust conditions can potentially result in many benefits, such as a steeper descent, a reduced landing run, reduced community noise, energy regeneration, etc. However, the aerodynamics and aeroacoustics of propellers in this regime are not well understood. This paper presents an aeroacoustic analysis of an isolated propeller operating in both positive and negative thrust conditions, using scale-resolved lattice-Boltzmann very large eddy simulations and the Ffowcs Williams & Hawkings analogy. The propeller was operated at a constant tip Mach number so that any differences in tonal noise between positive and negative thrust conditions were due to changes in blade loading. Results showed that the flow separation around the blades in the negative thrust case led to a 2 to 6 times higher standard deviation in integrated thrust compared to the positive thrust case. The blade loading in the negative thrust case shows the amplitude of fluctuations up to 18% for inboard sections and up to 30% near the blade tip compared to the time-averaged loads. The noise in the propeller plane is 10 dB higher in the positive thrust regime than in the negative thrust regime at a given absolute thrust level of
. The lower noise at negative thrust is caused by two factors: the lower magnitude of the negative torque compared to the positive torque at a given thrust level and the shift of the blade loading inboard in the negative thrust condition due to the stall of the blade tip. Along the propeller axis, the negative thrust regime has 13-15 dB higher noise because of the increased broadband noise generated by the flow separation. In the negative thrust case, the noise along the propeller axis (89 dB) and propeller plane (92 dB) are comparable. However, this is not the case for the propulsive case. The comparison of noise in the vicinity of the propeller plane showed that using the propellers in negative thrust conditions allows for a steeper and quieter descent compared to a conventional descent; as long as the magnitude of the negative torque produced is equal to or less than the torque required to operate the propeller in a conventional landing
A Deep Learning Emotion Classification Framework for Low Resource Languages
No full textEmotion classification from text is the process of identifying and classifying emotions expressed in textual data. Emotions can be feelings such as anger, joy, suspense, sadness and neutral. Developing a machine learning model to identify emotions in a low-resourced language with a limited set of linguistic resources and annotated corpora is a challenge. This research proposes a Deep Learning Emotion Classification Framework to identify and classify emotions in low-resourced languages such as Hindi. The proposed framework combines a classification model and a low resource optimization technique in a novel way. An annotated corpus of Hindi short stories consisting of 20,304 sentences is used to train the models for predicting five categories of emotions: anger, joy, suspense, sadness, and neutral talk. To resolve the class imbalance in the dataset SMOTE technique is applied. The optimal classification model is selected through experimentation that compares machine learning models and pre-trained models. Machine learning and deep learning models are SVM, Logistic Regression, Random Forest, CNN, BiLSTM, and CNN+BiLSTM. The pre-trained models, mBERT, IndicBERT, and a hybrid model, mBERT+BiLSTM. The models are evaluated based on macro average recall, macro average precision, and macro average F1 score. Results demonstrate that the hybrid model mBERT+BiLSTM out perform other models with a test accuracy of 57%
Inferring object states and articulation modes from egocentric videos
Get PDFWe develop algorithms for understanding objects from the point of view of interacting with them. There are two key aspects to obtaining such an understanding. First, objects can occur in different states and we need features that are sensitive to such states. Second, different objects can be articulated in different ways and we need to understand how to correctly infer their modes of articulation. We propose self and weakly supervised techniques to obtain such an understanding of objects purely through observation of how humans interact with the world around them through their hands. Our experiments on the challenging EPIC- KITCHENS dataset show the merits of using human hands as a probe for understanding objects.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2023-05-01The student, Rishabh Goyal, accepted the attached license on 2021-04-27 at 12:26.The student, Rishabh Goyal, submitted this Thesis for approval on 2021-04-27 at 13:39.This Thesis was approved for publication on 2021-04-28 at 09:43.DSpace SAF Submission Ingestion Package generated from Vireo submission #16585 on 2021-09-16 at 17:06:08Made available in DSpace on 2021-09-17T02:34:49Z (GMT). No. of bitstreams: 2
GOYAL-THESIS-2021.pdf: 26378428 bytes, checksum: 5fb34e7c3f1a83a84f3c5ba07fefd80f (MD5)
LICENSE.txt: 4210 bytes, checksum: 37522f5c1ac72c7fb0aa7fa78a307c50 (MD5)
Previous issue date: 2021-04-28Embargo set by: Seth Robbins for item 118591
Lift date: 2023-09-17T02:34:57Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemAuthor requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Onl
Aerodynamics and Far-field Noise Emissions of a Propeller in Positive and Negative Thrust Regimes at Non-zero Angles of Attack
Get PDFThis paper studies the effect of operation at non-zero angles of attack on the aerodynamic performance and far-field noise emissions of an isolated propeller operating at positive and negative thrust conditions. To achieve this, scale-resolved lattice-Boltzmann very large eddy simulations coupled with the Ffowcs Williams & Hawkings analogy have been used. The results show that when the propeller operates with a 10◦ angle of attack at the positive thrust condition, the blade loading increases on the advancing side and decreases on the retreating side, leading to a 9.6% increase in integrated thrust (when computed along the propeller axis) and a negligible increase (0.1%) in propeller efficiency. Conversely, at the negative thrust condition, the operation at 10 deg angle of attack results in a 7.9% decrease in thrust magnitude and an 11.1% reduction in energy-harvesting efficiency. In this condition, the positively cambered blade sections exhibit dynamic stall at the 10◦ angle of attack, resulting in broadband fluctuations of up to 10% of the mean loading. As a result of the opposite change in absolute blade loading in the negative thrust condition compared to the positive thrust condition at the 10◦ angle of attack, the change in the noise directivity is also the opposite. Whereas in the positive thrust case, the noise increases in the region from which the propeller is tilted away (i.e., below the propeller at a positive angle of attack), in the negative thrust case, it is the other way around. This study highlights the need to account for non-zero angles of attack in propeller design and optimization analyses
Small world: Narrow, wide, and long replication of Goyal, van der Leij and Moraga‐Gonzélez (JPE 2006) and a comparison of EconLit and Scopus
Get PDFI undertake a narrow, wide, and long replication of Goyal, van der Leij and Moraga‐Gonzélez (2006, https://doi.org/10.1086/500990). Using social network analysis, they show that the Economics profession gradually evolved into a small world. Small worlds (or small world networks) have unique information transmission capabilities. The trend is explained by the emergence of frequently publishing researchers with many distinct co‐authors. In a social network, they resemble stars. The original results are robust to the usage of (I) another software, (II) a recent version of the originally used data, and (III) another database and a more sophisticated author disambiguation
Small World: Narrow, Wide and Long replication of Goyal, van der Leij and Moraga-González (JPE 2006) and a Comparison of EconLit and Scopus
No full textI undertake a narrow, wide and long replication of Goyal, van der Leij and Moraga-González (Journal of Political Economy 2006; 114(2): 403–412). Using social network analysis they show that the Economics profession gradually evolved into a small world. Small worlds (or small world networks) have unique information transmission capabilities. The trend is explained by the emergence of frequently publishing researchers with many distinct co-authors. In a social network they resemble stars. The original results are robust to the usage of (I) another software, (II) a recent version of the originally used data, and (III) another database and a more sophisticated author disambiguation
Small is the new big: An overview of newer supraglottic airways for children
No full textAlmost all supraglottic airways (SGAs) are now available in pediatric sizes. The availability of these smaller sizes, especially in the last five years has brought a marked change in the whole approach to airway management in children. SGAs are now used for laparoscopic surgeries, head and neck surgeries, remote anesthesia; and for ventilation during resuscitation. A large number of reports have described the use of SGAs in difficult airway situations, either as a primary or a rescue airway. Despite this expanded usage, there remains little evidence to support its usage in prolonged surgeries and in the intensive care unit. This article presents an overview of the current options available, suitability of one over the other and reviews the published data relating to each device. In this review, the author also addresses some of the general concerns regarding the use of SGAs and explores newer roles of their use in children
- …
