YouTube's Algorithm, Scientifically Explained скачать в хорошем качестве

YouTube's Algorithm, Scientifically Explained

I explain how YouTube chooses recommendations from billions of videos. If you've ever wondered, "how does the YouTube algorithm work behind the scenes?", this video is for you. 0:00 Introduction 0:32 YouTube's objectives 0:52 Example objective: Maximizing Watch Time 5:14 Determining a video's topics 9:07 Tuning the algorithm's knobs 11:20 Trading off different objectives 13:17 Fixing YouTube's biases 14:10 Recap and conclusion Key References: Deep Neural Networks for YouTube Recommendations (Covington, Adams, Sargin, RecSys 2016) Recommending what video to watch next: a multitask ranking system (Zhao et al., RecSys 2019) Q&R: A Two-Stage Approach toward Interactive Recommendation (Christakopoulou et al., KDD 2018) Andrew Ng's Video on "Why Deep Representations?"    • Why Deep Representations? (C1W4L04)   Wide and Deep Learning: https://ai.googleblog.com/2016/06/wid... #machinelearning #datascience #youtube #algorithm Stock footage: Pexels Music: Mixkit Animation uses the Manim library, an open source Python package created by 3Blue1Brown.