Train and Deploy Machine Learning Model With Web Interface - PyTorch & Flask

Github: https://github.com/imadelh/ML-web-app

Article: https://imadelhanafi.com/posts/train_deploy_ml_model/

Example: https://ml-app.imadelhanafi.com/

Finally Kaggle is doing this!

Course 4 of the deeplearning.ai TensorFlow Specialization is now available on Coursera! Combine everything you’ve learned to build a sophisticated sunspot prediction model using real-world data. Enroll now: https://www.coursera.org/specializations/tensorflow-in-practice

Multiple Post-docs at U. of Toronto in AI / ML / Deep Learning

Over the next 1-2 months, I expect to have multiple post-doc openings for different AI / ML / Deep Learning research projects at the U. of Toronto:

(1) Bayesian Methods for Deep Learning
(2) Deep Learning for Traffic Prediction and Deep RL for Traffic Control
(3) Deep Learning and Embedding Methods for Financial Prediction from Social Media
(4) Deep Learning for Conversational Recommendation Systems

Salaries for all positions would be approximately CAD $54,000 per year. Initial offers would be made for one year with possibility of renewal for one or more years. Post-docs are expected to publish in the top venues relevant to the project research area and to produce high-quality deliverable code for use by research funding partners.

If you are interested, please email [email protected] with the following:

(a) your CV (clearly listing all publications),
(b) your github or bitbucket public account link with projects that I can browse (if you don't have this, please do not apply),
(c) 1-2 sentences in your email stating which position(s) interest you and why you think you are appropriate for the position.

Deadline: rolling until all positions become available and are filled (all offers expected to be made by Oct 1, 2019 at the latest).

Dr. Scott P. Sanner
Assistant Professor, Industrial Engineering
Cross-appointed, Computer Science
Faculty Affiliate, Vector Institute
University of Toronto, Toronto, ON, Canada
Email: [email protected]
Website: https://d3m.mie.utoronto.ca/

PhD position on "Robust Deep Learning in the Physical World" at Bosch Center for Artificial Intelligence

PhD - Robust Deep Learning in the Physical World
Jan Hendrik Metzen (https://scholar.google.de/citations?user=w047VfEAAAAJ)
Bosch Center for Artificial Intelligence (bosch-ai.com)
Renningen, Germany (https://www.bosch.de/en/our-company/bosch-in-germany/renningen/)
Job description:
Deep learning (DL) has achieved remarkable results for perceptual tasks within the last decade. However, DL-based perception often lacks sufficient robustness for real-world applications, as exemplified by the existence of adversarial examples and the fragility in face of natural distortions not foreseen during training. Besides, there is growing evidence that DL-based perception works differently than human perception on a fundamental level, e.g. relying overly strong on texture cues and on brittle characteristics of the training data.

In this PhD, we want to work on fundamentally new methods for DL, for instance new network architectures, new training procedures, or new regularization schemes.
The results should be published at the top-tier machine learning venues.

Qualifications:
* Personality: Communicative and team player
* Working Practice: Independent, motivated to work in an interdisciplinary and international team
* Experience and Knowledge: With deep learning frameworks (TensorFlow, PyTorch, etc.), basic knowledge of machine learning and deep learning, strong programming skills, in particular Python and strong mathematical background
* Languages: Very good in English (written and spoken)
* Education: Excellent degree (Master) in computer science, mathematics or related fields with excellent marks.

Additional Information:
Please refer to https://smrtr.io/3jL_w for further information and how to apply.

Looking forward to your application!
Jan Hendrik Metzen

PhD Position in Artificial Intelligence and Robotics at Graz University of Technology


1 position of university assistant (research associate) for 4 years, 40 hours/week, starting 1 November 2019, at the Institute of Software Technology. The position is related to research and teaching in the area of application of artificial intelligence techniques in smart and flexible production.

Admission requirements:

Master's or diploma degree in computer science, computer engineering or software development.


expected qualifications:

* Basic knowledge and practical experience in one or more of the following areas:

+ Declarative Problem Solving
+ Knowledge Representation and Reasoning
+ Combinatorial Search and Optimization
+ Constraints and Preferences
+ Computational Complexity
+ Planning, Scheduling and Execution Monitoring
+ Multi-agent Systems
+ Data Mining and Machine Learning
+ Algorithms and Implementation

* social, communicative and didactic competence

* good academic performance

* good English language skills

Classification:

B 1 under the collective agreement for university employees; the monthly minimum remuneration for this use is currently € 2,864.50 gross (14 times per year) and may increase on the basis of the provisions of the collective agreement as a result of the crediting of previous experience specific to the job and other remuneration components linked to the specific features of the job.

Applications, curricula vitae and other documents must be sent to [email protected], preferably electronically, with a precise description of the position or the reference number, and must be received by the end of the application period at the latest.


Application deadline: 30 August 2019

Excited to announce the call for submissions to our workshop at NeurIPS 2019: "Tackling Climate Change with Machine Learning," Dec 13/14 in Vancouver, Canada.

Submission deadline: Sept 11
Details: https://www.climatechange.ai/NeurIPS2019_workshop

Behind the Selection of the NeurIPS 2019 Workshops"

By Neural Information Processing Systems Conference: https://medium.com/@NeurIPSConf/2019workshops-ec820e4d558e

#MachineLearning #Neurips2019 #DeepLearning #ReinforcementLearning #Neurips

Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
Chelsea Finn, Pieter Abbeel, Sergey Levine : https://arxiv.org/abs/1703.03400
#MachineLearning #ArtificialIntelligence #EvolutionaryComputing

Tensorflow implementation of U-GAT-IT

Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation

GitHub, by Junho Kim : https://github.com/taki0112/UGATIT

#tensorflow #unsupervisedlearning #generativemodels

A Tensorflow 2.0 library for deep learning model interpretability
Blog by Raphaël Meudec : https://blog.sicara.com/tf-explain-interpretability-tensorflow-2-9438b5846e35
#MachineLearning #DeepLearning #TensorFlow #Interpretability

An artificial neural network called “EmoNet” could recognize which emotions, out of 20 different categories, a human would feel in response to an image, challenging the prevailing view that emotions are independent from the sensory environment.

Read the research in our open-access journal, Science Advances: https://fcld.ly/x5rbro1

Emotion schemas are embedded in the human visual system
https://advances.sciencemag.org/content/advances/5/7/eaaw4358.full.pdf

If you didn't know, there is a collection of datasets, ready to use with TF
https://github.com/tensorflow/datasets

ImageNet-trained deep neural network exhibits illusion-like response to the Scintillating Grid

https://arxiv.org/abs/1907.09019v1

Deep Non-Rigid Structure from Motion. arxiv.org/abs/1907.13123

Multi-Frame Cross-Entropy Training for Convolutional Neural Networks in Speech Recognition. arxiv.org/abs/1907.13121

Improved mutual information measure for classification and community detection. arxiv.org/abs/1907.12581