AI

所属课程: 人工智能大变革(16星期,32课时)

*eremy

13岁半

发布于:29天前
浏览数:1

1. I think the best method is to start in the middle and then go in a random direction. If the loss decreases, go in that same direction. If it increases, go in the opposite direction. Repeat this process until the minimum loss is reached.

2. If i start in the rght direction, I am usually able to get it within 12 trie*, *ut if i start in the wrong direction, it will be much harder for me to get to the minimum loss.

3. This challenge helped me understand gradient decent better because it walked me through the actual process of finding the global minimum.

你还没有登录,请先登录注册
  • 还没有人评论,欢迎说说您的想法!