DEEP LEARNING FOR TIME SERIES

An hybrid event in Matam Haifa and online about time series:

Session #1: Time Series Classification Algorithms Review (Tamir Nave, AI-Blog, ALMA)

In this lecture Tamir will go through some selected algorithms from recent years that solve the time series classification problem. The algorithms that will be covered are: shapelet, BOSS (bag of SFA symbols), Proximity Forest, InceptionTime and Rocket (Random Convolutional Kernal Transform). The purpose of the lecture is to give the main idea of each method briefly without getting into implementation details.

Session #2: Koopman operator for high-dimensional time series prediction (Barr Morgenstein, Autobrain and She Codes)

In this talk, Barr will introduce the She Codes program and their initiatives to advance women in the high-tech world, followed by an introduction to the Koopman operator as a tool in unsupervised learning. Barr will provide the background and use cases on which the operator enables the extraction of key features from high-dimensional dynamic data.

Session #3: Off Road Autonomous Grading (Yakov Miron, Bosch)

In this talk, Yakov will explain about the autonomous grading task for off road autonomous bulldozers.
He will point the main challenges, sensors that are used, how BCAI tackle the problem and will elaborate on their approach to train an RL agent from scratch.

Session #4: KalmanNet: Model-based deep learning. Deep learning augmented Kalman filtering and State Space Models (Guy Revach,ETH Zürich)

In this talk, Guy will introduce model-based deep learning and more specifically combining deep learning and state-space models for the task of Kalman filtering and Kalman smoothing

Who can join us?
This session is for people with advanced background in the field of machine learning. Some of the content will be technical and academic-oriented.

Attend the event in Matam Haifa requires registration at: oria@almatechnologies.com.

All talks will be given in english.

Schedule:
18:00 – Welcoming – Introduce by Barak Or
18:20 - Lecture #1 – “Time Series Classification Algorithms Review” by Tamir Nave, AI-Blog, ALMA
18:40 - Lecture #2 – “Koopman operator for high-dimensional time series prediction” by Barr Morgenstein, Autobrain and She Codes
19:00 – Lecture #3 - “Off Road Autonomous Grading” by Yakov Miron, Bosch
19:20 - Lecture #4 - “KalmanNet: Model-based deep learning. Deep learning augmented Kalman filtering and State Space Models” by Guy Revach,ETH Zürich

About AI-BLOG:
AI-BLOG, the first and leading Hebrew blog on machine learning, deep learning, data science, and AI.
AI-Blog Community is about people that help each other to become more professional. www.ai-blog.co.il
About the Speakers:

Barak Or is the founder & CEO at ALMA. Artificial intelligence and sensor fusion expert; Ex-Qualcomm, entrepreneur. Barak is a Ph.D. candidate and holds MSc and BSc in engineering, and BA in Economics from the Technion. He has authored several patents and articles that have been published in professional journals. Winner of Gemunder prize.

Yakov Miron is a research Engineer at the Bosch Center for Artificial Intelligence (BCAI), working on Computer Vision (CV) and Reinforcement Learning (RL). Currently, his main research interest is the interaction of CV and RL. Specifically, working on off road autonomous driving and path planning.

Barr Morgenstein is an Algorithm Developer at Autobrains, Python and Data Analysis Supervisor at SheCodes.
Bar holds M.Sc. in Electrical Engineering from NYU Tandon School of Engineering and B.Sc. in Electrical Engineering from Tel-Aviv University.

Guy Revach holds a BSc and MSc from the Electrical engineering department at the Technion. Guy is currently a Ph.D. candidate in the Institute for Signal and Information Processing at ETH Zürich, Switzerland.

Tamir Nave has 10 years experience in the field of algorithms, focusing on computer vision, deep learning and reinforcement learning as a developer, team leader, consultant and mentor.