Search

Your search keyword '"Nichol, Alan"' showing total 504 results
Start Over Author "Nichol, Alan"
504 results on '"Nichol, Alan"'

1. Task-Oriented Dialogue with In-Context Learning

Author

Bocklisch, Tom, Werkmeister, Thomas, Varshneya, Daksh, and Nichol, Alan

Subjects
Computer Science - Computation and Language
Abstract

We describe a system for building task-oriented dialogue systems combining the in-context learning abilities of large language models (LLMs) with the deterministic execution of business logic. LLMs are used to translate between the surface form of the conversation and a domain-specific language (DSL) which is used to progress the business logic. We compare our approach to the intent-based NLU approach predominantly used in industry today. Our experiments show that developing chatbots with our system requires significantly less effort than established approaches, that these chatbots can successfully navigate complex dialogues which are extremely challenging for NLU-based systems, and that our system has desirable properties for scaling task-oriented dialogue systems to a large number of tasks. We make our implementation available for use and further study.

Published
2024

5. Where is the context? -- A critique of recent dialogue datasets

Author

Mosig, Johannes E. M., Vlasov, Vladimir, and Nichol, Alan

Subjects
Computer Science - Computation and Language
Abstract

Recent dialogue datasets like MultiWOZ 2.1 and Taskmaster-1 constitute some of the most challenging tasks for present-day dialogue models and, therefore, are widely used for system evaluation. We identify several issues with the above-mentioned datasets, such as history independence, strong knowledge base dependence, and ambiguous system responses. Finally, we outline key desiderata for future datasets that we believe would be more suitable for the construction of conversational artificial intelligence.

Published
2020

6. DIET: Lightweight Language Understanding for Dialogue Systems

Author

Bunk, Tanja, Varshneya, Daksh, Vlasov, Vladimir, and Nichol, Alan

Subjects
Computer Science - Computation and Language
Abstract

Large-scale pre-trained language models have shown impressive results on language understanding benchmarks like GLUE and SuperGLUE, improving considerably over other pre-training methods like distributed representations (GloVe) and purely supervised approaches. We introduce the Dual Intent and Entity Transformer (DIET) architecture, and study the effectiveness of different pre-trained representations on intent and entity prediction, two common dialogue language understanding tasks. DIET advances the state of the art on a complex multi-domain NLU dataset and achieves similarly high performance on other simpler datasets. Surprisingly, we show that there is no clear benefit to using large pre-trained models for this task, and in fact DIET improves upon the current state of the art even in a purely supervised setup without any pre-trained embeddings. Our best performing model outperforms fine-tuning BERT and is about six times faster to train., Comment: v3: Updated results for the best model

Published
2020

9. Dialogue Transformers

Author

Vlasov, Vladimir, Mosig, Johannes E. M., and Nichol, Alan

Subjects
Computer Science - Computation and Language
Abstract

We introduce a dialogue policy based on a transformer architecture, where the self-attention mechanism operates over the sequence of dialogue turns. Recent work has used hierarchical recurrent neural networks to encode multiple utterances in a dialogue context, but we argue that a pure self-attention mechanism is more suitable. By default, an RNN assumes that every item in a sequence is relevant for producing an encoding of the full sequence, but a single conversation can consist of multiple overlapping discourse segments as speakers interleave multiple topics. A transformer picks which turns to include in its encoding of the current dialogue state, and is naturally suited to selectively ignoring or attending to dialogue history. We compare the performance of the Transformer Embedding Dialogue (TED) policy to an LSTM and to the REDP, which was specifically designed to overcome this limitation of RNNs., Comment: 10 pages, 4 figures, 1 table

Published
2019

12. Few-Shot Generalization Across Dialogue Tasks

Author

Vlasov, Vladimir, Drissner-Schmid, Akela, and Nichol, Alan

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

Machine-learning based dialogue managers are able to learn complex behaviors in order to complete a task, but it is not straightforward to extend their capabilities to new domains. We investigate different policies' ability to handle uncooperative user behavior, and how well expertise in completing one task (such as restaurant reservations) can be reapplied when learning a new one (e.g. booking a hotel). We introduce the Recurrent Embedding Dialogue Policy (REDP), which embeds system actions and dialogue states in the same vector space. REDP contains a memory component and attention mechanism based on a modified Neural Turing Machine, and significantly outperforms a baseline LSTM classifier on this task. We also show that both our architecture and baseline solve the bAbI dialogue task, achieving 100% test accuracy.

Published
2018

13. Dose-painted volumetric modulated arc therapy of high-grade glioma using 3,4-dihydroxy-6-[18F]fluoro-L-phenylalanine positron emission tomography

Author

Kosztyla, Robert, Raman, Srinivas, Moiseenko, Vitali, Reinsberg, Stefan A, Toyota, Brian, and Nichol, Alan

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Oncology and Carcinogenesis, Bioengineering, Biomedical Imaging, Rare Diseases, Cancer, Adult, Aged, Aged, 80 and over, Brain, Brain Neoplasms, Dihydroxyphenylalanine, Female, Glioma, Humans, Male, Middle Aged, Organs at Risk, Positron-Emission Tomography, Radiotherapy Dosage, Radiotherapy Planning, Computer-Assisted, Radiotherapy, Intensity-Modulated, Retrospective Studies, Tumor Burden, Young Adult, Nuclear Medicine & Medical Imaging, Clinical sciences, Oncology and carcinogenesis
Abstract

ObjectiveTo determine whether dose painting with volumetric modulated arc therapy for high-grade gliomas using 3,4-dihydroxy-6-[18F]fluoro-l-phenylalanine (18F-FDOPA) positron emission tomography (PET) could achieve dose-escalated coverage of biological target volumes (BTVs) without increasing the dose to cranial organs at risk (OARs).Methods10 patients with high-grade gliomas underwent CT, MRI, and 18F-FDOPA PET/CT images for post-operative radiation therapy planning. Two volumetric modulated arc therapy plans were retrospectively generated for each patient: a conventional plan with 60 Gy in 30 fractions to the planning target volume delineated on MRI and a dose-escalated plan with a maximum dose of 80 Gy in 30 fractions to BTVs. BTVs were created by thresholding 18F-FDOPA PET/CT uptake using a linear quadratic model that assumed tracer uptake was linearly related to tumour cell density. The maximum doses and equivalent uniform doses of OARs were compared.ResultsThe median volume of the planning target volume receiving at least 95% of the prescribed dose (V 95%) was 99.6% with and 99.5% without dose painting. The median V 95% was >99.2% for BTVs. The maximum doses and equivalent uniform doses to the OARs did not differ significantly between the conventional and dose-painted plans.ConclusionUsing commercially available treatment planning software, dose painting for high-grade gliomas was feasible with good BTV coverage and no significant change in the dose to OARs.Advances in knowledgeA novel treatment planning strategy was used to achieve dose painting for gliomas with BTVs obtained from 18F-FDOPA PET/CT using a radiobiological model.

Published
2019

14. Intra-individual comparison of 18F-sodium fluoride PET–CT and 99mTc bone scintigraphy with SPECT in patients with prostate cancer or breast cancer at high risk for skeletal metastases (MITNEC-A1): a multicentre, phase 3 trial

Author

Bénard, François, Harsini, Sara, Wilson, Don, Zukotynski, Katherine, Abikhzer, Gad, Turcotte, Eric, Cossette, Mariève, Metser, Ur, Romsa, Jonathan, Martin, Montgomery, Mar, Colin, Saad, Fred, Soucy, Jean-Paul, Eigl, Bernhard J, Black, Peter, Krauze, Andra, Burrell, Steven, Nichol, Alan, and Tardif, Jean-Claude

Published
2022
Full Text
View/download PDF

17. Rasa: Open Source Language Understanding and Dialogue Management

Author

Bocklisch, Tom, Faulkner, Joey, Pawlowski, Nick, and Nichol, Alan

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence, Computer Science - Learning
Abstract

We introduce a pair of tools, Rasa NLU and Rasa Core, which are open source python libraries for building conversational software. Their purpose is to make machine-learning based dialogue management and language understanding accessible to non-specialist software developers. In terms of design philosophy, we aim for ease of use, and bootstrapping from minimal (or no) initial training data. Both packages are extensively documented and ship with a comprehensive suite of tests. The code is available at https://github.com/RasaHQ/, Comment: Presented at NIPS Workshop on Conversational AI, Code at https://github.com/RasaHQ

Published
2017

20. General Techniques for Radiosurgery

Author

Ruschin, Mark, Sahgal, Arjun, Ma, Lijun, Wang, Lei, Gete, Ermias, Nichol, Alan, Yamada, Yoshiya, editor, Chang, Eric, editor, Fiveash, John B., editor, and Knisely, Jonathan, editor

Published
2020
Full Text
View/download PDF

23. Relating melting trends and elasticity in simple metals: an empirical potential approach

Author

Nichol, Alan and Ackland, Graeme J.

Subjects
Condensed Matter - Materials Science
Abstract

We demonstrate that the melting points and other thermodynamic quantities of the alkali metals can be calculated based on static crystalline properties. To do this we derive analytic interatomic potentials for the alkali metals fitted precisely to cohesive and vacancy energies, elastic moduli, lattice parameter and crystal stability. These potentials are then used to calculate melting points by simulating the equilibration of solid and liquid samples in thermal contact at ambient pressure. With the exception of lithium, remarkably good agreement is found with experimental values. The instability of the bcc structure in Li and Na at low temperatures is also reproduced, and, unusually, is not due to a soft T1N phonon mode. No forces or finite temperature properties are included in the fit, so this demonstrates a surprisingly high level of intrinsic transferrability in the simple potentials. Currently, there are few potentials available for the alkali metals, so in, addition to demonstrating trends in behaviour, we expect that the potentials will be of broad general use.

Published
2016
Full Text
View/download PDF

25. The use of bolus in postmastectomy radiation therapy for breast cancer: A systematic review

Author

Dahn, Hannah M., Boersma, Liesbeth J., de Ruysscher, Dirk, Meattini, Icro, Offersen, Birgitte V., Pignol, Jean-Philippe, Aristei, Cynthia, Belkacemi, Yazid, Benjamin, Dori, Bese, Nuran, Coles, Charlotte E., Franco, Pierfrancesco, Ho, Alice, Hol, Sandra, Jagsi, Reshma, Kirby, Anna M., Marrazzo, Livia, Marta, Gustavo N., Moran, Meena S., Nichol, Alan M., Nissen, Henrik D., Strnad, Vratislav, Zissiadis, Yvonne E., Poortmans, Philip, and Kaidar-Person, Orit

Published
2021
Full Text
View/download PDF

28. Quality of life after breast-conserving therapy and adjuvant radiotherapy for non-low-risk ductal carcinoma in situ (BIG 3-07/TROG 07.01): 2-year results of a randomised, controlled, phase 3 trial

Author

Chua, Boon H, Phillips, Claire, Bryant, Guy, Westenberg, Helen, Purohit, Om Pra-Kesh, Ahern, Verity, Graham, Peter, Akra, Mohamed, McArdle, Orla, O'Brien, Peter, Ludbrook, Jane, Harvey, Jennifer, Maduro, John H, Gabelle-Flandin, Isabelle, Kirkove, Carine, Bedi, Carolyn, Martin, Joseph, Vu, Tony, Muanza, Theirry, Neal, Anthony, Courdi, Adel, Thariat, Juliette, Rakovitch, Eileen, Daniels, Laurien, van Hezewijk, Marjan, Cwajna, Wlasyslawa, Roelstraete, Adelheid, van Baardwijk, Angela, Russel, Nicola, Koch, Anne, Croke, Jennifer, Locke, Imogen, Jeal, Peter, Walker, Quenten, Thuraisingham, Kandeepeepan, Chauduri, Anupam, Joseph, David, Taylor, Mandy, Vanderkam, Sabine, Woo, Tony, Tang, Johann, Yassa, Michael, Wai, Elaine, Hewitt, Susan, Mahmood, Shazia, Gilmore, Jennifer, Ofi, Bolante, Bahl, Amit, Vujovic, Olga, Yu, Edward, Le, Duc, Kong, Iwa, Nichol, Alan, Bijker, Nina, Delaney, Geoff, Feigen, Malcolm, Lim, Adeline, Chao, Michael, Latham, Margaret, Algurafi, Hafiz, Tausch, Christoph, Khoo, Eric, Leung, Sam, Taylor, Karen, Senthi, Sasha, Stevens, Andrea, Chaudhuri, Abhro, Cleator, Susan, Brunt, Adrian Murray, Babington, Scott, Christie, David, Zwahlen, Daniel, Schratzenstaller, Ulrich, Masson, Laurence, Storey, Nicola, Kumar, Eshwar, Sherwin, Liz, Weytjens, Reinhilde, Ravi, Sharma, Lawton, Patricia, Angell, Ruth, Round, Glenys, Allen, Angela, Thotathil, Ziad, Anthes, Margaret, Reuter, Christiane, Pettit, Laura, Zissiadis, Yvonne, Elder, Christine, Verbeek-de Kanter, Antoinette, Lirette, Andree, Plasswilm, Ludwig, Spooner, David, Hoar, Fiona, Mohamed, Islam, Lossl, Kristina, Loo, Vivienne, Richetti, Antonella, Evans, Tamasin, Hennessy, Aisling, El-Mallah, Medhat, Skala, Marketa, Awad, Raef, Germain, Isabelle, Mitine, Carine, Van Parijs, Hilde, Churn, Mark, Walji, Nawaz, Francis, Michael, Stellamans, Karin, Gruber, Gunther, Ivaldi, Giovanni, Alhasso, Abdulla, Kenny, Lizbeth, Tiver, Ken, Griffin, Matthew, Lamoury, Gillian, Trovo, Marco, Algufarfi, Hafiz, Lah, Minjae, Carruthers, Scott, Papadatos, George, Paardekooper, Gabriel, Persic, Mojca, Lavery, Bernadette, King, Madeleine T, Link, Emma K, Whelan, Tim J, Olivotto, Ivo A, Kunkler, Ian, Westenberg, Antonia Helen, Gruber, Guenther, and Schofield, Penny

Published
2020
Full Text
View/download PDF

30. Interhemispheric Difference Images from Postoperative Diffusion Tensor Imaging of Gliomas

Author

Kosztyla, Robert, Reinsberg, Stefan A, Moiseenko, Vitali, Toyota, Brian, and Nichol, Alan

Subjects
Clinical Research, Neurosciences, Brain Cancer, Cancer, Biomedical Imaging, Brain Disorders, Rare Diseases, glioma, diffusion tensor imaging, magnetic resonance imaging, image analysis, target volumes, Medical and Health Sciences
Abstract

Introduction Determining the full extent of gliomas during radiotherapy planning can be challenging with conventional T1 and T2 magnetic resonance imaging (MRI). The purpose of this study was to develop a method to automatically calculate differences in the fractional anisotropy (FA) and mean diffusivity (MD) values in target volumes obtained with diffusion tensor imaging (DTI) by comparing with values from anatomically homologous voxels on the contralateral side of the brain. Methods Seven patients with a histologically confirmed glioma underwent postoperative radiotherapy planning with 1.5 T MRI and computed tomography. DTI was acquired using echo planar imaging for 20 noncolinear directions with b = 1000 s/mm2 and one additional image with b = 0, repeated four times for signal averaging. The distribution of FA and MD was calculated in the gross tumor volume (GTV), shells 0-5 mm, 5-10 mm, 10-15 mm, 15-20 mm, and 20-25 mm outside the GTV, and the GTV mirrored in the left-right direction (mirGTV). All images were aligned to a template image, and FA and MD interhemispheric difference images were calculated. The difference in mean FA and MD between the regions of interest was statistically tested using two-sided paired t-tests with α = 0.05. Results The mean FA in mirGTV was 0.20 ± 0.04, which was larger than the FA in the GTV (0.12 ± 0.03) and shells 0-5 mm (0.15 ± 0.03) and 5-10 mm (0.17 ± 0.03) outside the GTV. The mean MD (×10-3 mm2/s) in mirGTV was 0.93 ± 0.09, which was smaller than the MD in the GTV (1.48 ± 0.19) and the peritumoral shells. The distribution of FA and MD interhemispheric differences followed the same trends as FA and MD values. Conclusions This study successfully implemented a method for calculation of FA and MD differences by comparison of voxel values with anatomically homologous voxels on the contralateral side of the brain. Further research is warranted to determine if radiotherapy planning using these images can be used to improve target delineation.

Published
2016

34. Variable dose interplay effects across radiosurgical apparatus in treating multiple brain metastases

Author

Ma, Lijun, Nichol, Alan, Hossain, Sabbir, Wang, Brian, Petti, Paula, Vellani, Rosemin, Higby, Chris, Ahmad, Salahuddin, Barani, Igor, Shrieve, Dennis C, Larson, David A, and Sahgal, Arjun

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Brain Neoplasms, Humans, Neoplasm Metastasis, Radiosurgery, Radiotherapy Dosage, Radiotherapy Planning, Computer-Assisted, Stereotactic radiosurgery, Gamma Knife, Intensity modulation, Brain metastases, Nuclear Medicine & Medical Imaging, Clinical sciences, Computer vision and multimedia computation
Abstract

PurposeNormal brain tissue doses have been shown to be strongly apparatus dependent for multi-target stereotactic radiosurgery. In this study, we investigated whether inter-target dose interplay effects across contemporary radiosurgical treatment platforms are responsible for such an observation.MethodsFor the study, subsets ([Formula: see text] and 12) of a total of 12 targets were planned at six institutions. Treatment platforms included the (1) Gamma Knife Perfexion (PFX), (2) CyberKnife, (3) Novalis linear accelerator equipped with a 3.0-mm multi-leaf collimator (MLC), and the (4) Varian Truebeam flattening-filter-free (FFF) linear accelerator also equipped with a 2.5 mm MLC. Identical dose-volume constraints for the targets and critical structures were applied for each apparatus. All treatment plans were developed at individual centers, and the results were centrally analyzed.ResultsWe found that dose-volume constraints were satisfied by each apparatus with some differences noted in certain structures such as the lens. The peripheral normal brain tissue doses were lowest for the PFX and highest for TrueBeam FFF and CyberKnife treatment plans. Comparing the volumes of normal brain receiving 12 Gy, TrueBeam FFF, Novalis, and CyberKnife were 180-290% higher than PFX. The mean volume of normal brain-per target receiving 4-Gy increased by approximately 3.0 cc per target for TrueBeam, 2.7 cc per target for CyberKnife, 2.0 cc per target for Novalis, and 0.82 cc per target for PFX. The beam-on time was shortest with the TrueBeam FFF (e.g., 6-9 min at a machine output rate of 1,200 MU/min) and longest for the PFX (e.g., 50-150 mins at a machine output rate of 350 cGy/min).ConclusionThe volumes of normal brain receiving 4 and 12 Gy were higher, and increased more swiftly per target, for Linac-based SRS platforms than for PFX. Treatment times were shortest with TrueBeam FFF.

Published
2014

40. 23 A Retrospective Study Comparing Outcomes in Matched Patients with Poor Prognosis Treated with Targeted Conventional Volumetric Modulated ARC Radiotherapy Versus Whole Brain Radiotherapy for Brain Metastases

Author

Parmar, Gurjit, primary, Kong, Timothy, additional, Hamm, Jeremy, additional, Liu, Mitchell, additional, Lefresne, Shilo, additional, Carolan, Hannah, additional, Berthelet, Eric, additional, Chan, Jessica, additional, and Nichol, Alan, additional

Published
2023
Full Text
View/download PDF

41. 202 Safety and Efficacy of Fractionated Stereotactic Radiation Therapy for Pituitary Adenomas

Author

Ylanan, Aveline Marie, primary, Nichol, Alan, additional, Ma, Roy, additional, McKenzie, Michael, additional, Hsu, Fred, additional, Ahmed, Sheikh Nisar, additional, Cheung, Arthur, additional, Akagami, Ryojo, additional, Gooderham, Peter, additional, Johnson, Michelle, additional, Gete, Ermias, additional, Vallieres, Isabelle, additional, and Sharieff, Waseem, additional

Published
2023
Full Text
View/download PDF

42. 96 Breast Cancer Outcomes of Patients with T4B Disease: A Population-Based Retrospective Analysis096 Breast Cancer Outcomes of Patients with T4B Disease: A Population-Based Retrospective Analysis

Author

Siriani-Ayoub, Nicolas, primary, Koulis, Theodora, additional, Gondara, Lovedeep, additional, Speers, Caroline, additional, Diocee, Rekha, additional, Lohrisch, Caroline, additional, Nichol, Alan, additional, Truong, Pauline, additional, and Siriani-Ayoub, Nicolas, additional

Published
2023
Full Text
View/download PDF

44. A protocol for a systematic review with prospective individual patient data meta-analysis in EGFR mutant NSCLC with brain metastases to assess the effect of SRS + Osimertinib compared to Osimertinib alone: the STARLET collaboration

Author

Robledo, Kristy P, primary, Lefresne, Shilo, additional, Yang, Yu, additional, Sahgal, Arjun, additional, Pinkham, Mark B, additional, Nichol, Alan, additional, Soo, Ross Andrew, additional, Parmar, Ambika, additional, Hegi-Johnson, Fiona, additional, Doherty, Mark, additional, Solomon, Benjamin J, additional, Shultz, David, additional, Tham, Ivan WK, additional, Sacher, Adrian, additional, Tey, Jeremy, additional, Leong, Cheng Nang, additional, Koh, Wee Yao, additional, Huang, Yiqing, additional, Ang, Yvonne, additional, Low, Jiali, additional, Yong, Clement, additional, Lim, Mei Chin, additional, Tan, Ai Peng, additional, Lee, Chee Khoon, additional, and Ho, Cheryl, additional

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results