Predicting Prosodic Prominence from Text with Pre-trained Contextualized Word Representations

Talman et al.: https://arxiv.org/abs/1908.02262

GitHub: https://github.com/Helsinki-NLP/prosody

#dataset #machinelearning #naturallanguageprocessing

Structured Knowledge Discovery from Massive Text Corpus. arxiv.org/abs/1908.01837

Risk Management via Anomaly Circumvent: Mnemonic Deep Learning for Midterm Stock Prediction. arxiv.org/abs/1908.01112

Simultaneous Semantic Segmentation and Outlier Detection in Presence of Domain Shift arxiv.org/abs/1908.01098

U-Net Fixed-Point Quantization for Medical Image Segmentation. arxiv.org/abs/1908.01073

Postdoc and PhD positions in hydrologic machine learning at Penn State University


In anticipation of incoming vacancies, I am seeking both a postdoctoral scholar and a PhD graduate student to work on projects utilizing recent progress in deep learning in hydrologic and hydrology-related predictions. The scholars will have the chance to work on multi-disciplinary projects including the Google AI Impacts Challenge project (https://sites.google.com/view/deepldb) and a new project that examines the interactions between hydrology and other domains. A full description of both projects can be solicited from Dr. Chaopeng Shen ([email protected]) by email. The selected candidates will interact with large groups of researchers, will be using state-of-the-art machine learning techniques and will learn new methodologies in solving complex issues.

This is an informal invitation for interested applicants to contact Shen. The formal job application portal for the postdoc position will be open later. The postdoc position may start as soon as October this year or early next year. Strong, documented coding skill is a must for both positions, while it is not limited to a particular programming language. Background in machine learning or deep learning is a plus but not required.

--
Chaopeng Shen, Associate Professor
Department of Civil and Environmental Engineering
231C Sackett Building
The Pennsylvania State University
University Park, PA 16802
Email: [email protected]
Twitter: @ChaopengShen
Office: 814-863-5844
Lab website: https://water.engr.psu.edu/shen/

Fully funded PhD Scholarship - Machine learning & AI for activity recognition (2019)

University of Sussex.

Open to candidates from UK/EU countries and from non-EU countries.

Deadline: 16 August 2019 17:00 (GMT)

https://www.sussex.ac.uk/study/fees-funding/phd-funding/view/1091-Fully-funded-PhD-Scholarship-Machine-learning-AI-for-activity-recognition

U-GAT-IT

Official TensorFlow Implementation : https://github.com/taki0112/UGATIT

#DeepLearning #Tensorflow #UnsupervisedLearning

Postdoc position in Machine Learning @ German Aerospace Center (DLR)

The German Aerospace Center's Institute of Data Science in Jena, Germany, is currently seeking a postdoc with at least 3 years' experience in deep learning.

Our newly established machine learning group focuses on developing deep learning approaches for a broad range of applications. This group maintains a close cooperation with the Remote Sensing Technology Institute in Oberpfaffenhofen, and is in frequent contact with other institutes of the German Aerospace Center (DLR). As such, the developed methods will be applied and evaluated in various domains with close relevance to applications within DLR, particularly earth observation. Quality control for these methods is also emphasized; aspects of this include validation, robustness, uncertainty modeling, and interpretability. This position will focus in particular on the development of novel approaches for low-resource tasks and noisy data.

A full description of the position, including the link to the application portal, is available here:
https://www.dlr.de/dlr/jobs/en/desktopdefault.aspx/tabid-10596/1003_read-33291/

For questions, please do not hesitate to contact me.

--
Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR)
Institute of Data Science | Data Management and Analysis | Mälzerstraße 3 | D-07745 Jena
Dr. Anna Kruspe | Acting Group Lead Machine Learning
Telefon +49 3641 30960 127 | [email protected]
DLR.de

Predicting Prosodic Prominence from Text with Pre-trained Contextualized Word Representations

Talman et al.: https://arxiv.org/abs/1908.02262

GitHub: https://github.com/Helsinki-NLP/prosody

#dataset #machinelearning #naturallanguageprocessing

Example of how deep fake can save the world)
https://www.youtube.com/watch?v=Y1HGgICqZ3c

At the heart of most deep learning generalization bounds (VC, Rademacher, PAC-Bayes) is uniform convergence (u.c.). We argue why u. c. may be unable to provide a complete explanation of generalization, even if we take into account the implicit bias of SGD.

https://arxiv.org/pdf/1902.04742.pdf

https://t.iss.one/ArtificialIntelligenceArticles

Deep-learning algorithms can be applied to large datasets of electrocardiograms, are capable of identifying abnormal heart rhythms and mechanical dysfunction, and could aid healthcare decisions: (link: https://go.nature.com/2HFYd2o) go.nature.com/2HFYd2o

ice work on how to sync different videos
https://ai.googleblog.com/2019/08/video-understanding-using-temporal.html

Few things are closer to my heart than Graphs and Machine Learning. This survey paper looks at deep learning techniques for handling graphs:

https://arxiv.org/pdf/1812.04202.pdf

Stanford computer scientists developed a deep learning algorithm that can diagnose heart rhythm defects better than cardiologists.

https://stanford.io/2BnLYzB

Excellent post on achieving state-of-the-art heart disease diagnosis using deep learning (dilated U-Net in Keras): https://blog.insightdatascience.com/heart-disease-diagnosis-with-deep-learning-c2d92c27e730 https://t.iss.one/ArtificialIntelligenceArticles

Know here everything about Julia - the programming language making Machine Learning (ML) better. 💻

https://appinventiv.com/blog/introducing-julia-machine-learning-development-language/

DoorGym: A Scalable Door Opening Environment and Baseline Agent
Urakami et al.: https://arxiv.org/pdf/1908.01887v1.pdf
#DeepLearning #ReinforcementLearning #Robotics