Machine & Deep learning Israel highlights: 13.1-20.1.2018

מאת אורי אליאבייב, 22 בינואר 2019

כמידי שבוע אני מאגד לכם את כל הפוסטים הדיונים החשובים שעלו בקבוצה. השבוע היו לא מעט דיונים מרתקים על מגוון רחב של נושאים שכדאי לכם לבדוק. כפי שאתם בוודאי יודעים, הרשימה הזו נשלחת גם בניוזלטר שלנו כמידי שבוע (הירשמו כאן!). את הניוזלטר המלא ניתן למצוא כאן (השבוע: הקורס הרביעי של Razor, מגוון עבודות חדשות ועוד).


 Group's Highlights from last week

We had a great week and many good posts and links. Let's start.

1. Gidi Shperber has shared an article about "Data Science and the Art of Persuasion" which talks about why sometimes Data scientists can't meet the expectations of their organization. This is a very important topic which is very close to my heart and an important dissociation we all should have. If you need to read only one link, make it this one.

2. Gad Benram shared a dataset of the "AutoTel Cars Locations" that his company, DoiT International, has gathered. Go and check out this database, you could do interesting stuff with it.

3. Aviv Rotman asked what is your favourite clustering algorithm and why it's better than kmeans. He got some wonderful answers.

4. Yoav Ramon wanted to know how others search for the best hyperparameters.

5. Yam Peleg shared many Reinforcement Learning algorithms written at Pytorch. You might need it someday :).

6. Amitai Armon shared a link of an article about AI and local politics.

7. Shlomo Kashani: "What happens when a researcher (Ren Li) decides to question each and every claim made in a "revolutionary" deep learning method (Spampinato et al.) for the analysis of EEG signals?". That's what happens.

8. Uri Berkovitz from Globes shared his first article at the group: An interview with Pro. Naftali Tishby. Some didn't agree with all the parts in the article.

Dan Ofer shared a link to this article: Using the latest advancements in deep learning to predict stock price movements". You can learn many good new tricks.

9. The 10-year challenge – the conspiracy.

ככעגככ כגעגכע