ML Lecture 0-1: Introduction of Machine Learning

[cryptoinside]

Mr. Lee is awesome....his presentation is so structured and logical. Appreciate the efforts putting into this.

[HandledHandleog]

感謝老師！以下幾個點讓我覺得學習到了：
26:27 Transfer Learning: data not related to the task considered
27:36 Unsupervised Learning-Machine Reading: input:a word, output: a vector representing the attributes of the word.
29:13 Structured Learning: output: structured output(speech recognition, machine translation..) Technique: GAN
32:32 Supervised vs. Reinforcement: learning for teacher vs. critics
36:02 Relation between Terminology

[StevenKing-sh]

阿老师， 我从哔哩哔哩网站追过来了，没想到你在youtube 也有开课， 听说你是机器学习课程的大牛！

[smartbodhi]

最近想要进入机器学习领域，来 youtube 搜索相关课程，发现这个，谢谢李老师无私奉献。

[goodgoodstudy]

谢谢老师！ 以下知识点我觉得学习到了：

**Supervised Learning**

1. Regression Problem: Output is “a scalar”
2. Classification Problem: Output is “class 1, class 2, , class N”
1. Linear Model
2. Non-Linear Model: Deep Learning / SVM / KNN etc..
3. Structure Learning Problem: Output is “a sequence/matrix/graph/tree” etc..

pay attention:

1. You need so much training data, means you should have input/output pair of target function and the function output also named the label.
2. Also, it’s hard to collect a larget amount of labelled data.

**Semi-Supervised Learning & Transfer Learning**

for example, recognizing the cats and dogs.

1. for semi-supervised learning, now you only have
1. little labelled data recognizied the cat and dog
2. much unlabelled data haven’t recognizied

surprisingly, the semi-supervised learning could learn from these data.

2. for transfer learning, now you only have
1. little labelled data recognizied the cat and dog
2. much data not related to the task considered (can be labeled or unlabeled)

surprisingly, the transfer learning learnt from these data to lookup whether these unrelated data is helpful


**UnSupervised Learning**

Unsupervised Learning is a machine learning technique in which the users do not need to supervise the model. Instead, it allows the model to work on its own to discover patterns and information that was previously undetected. It mainly deals with the unlabelled data.

**Reinforcement Learning**

Reinforcement learning is a machine learning training method based on rewarding desired behaviors and/or punishing undesired ones. In general, a reinforcement learning agent is able to perceive and interpret its environment, take actions and learn through trial and error. In one word, the reinforcement learning is learnt from critics. Alpha go is supervised learning + reinforcement learning.


🥰🥰🥰🥰🥰🥰🥰🥰

[cryptoinside]

“whenever someone says ‘AI’ what they're really talking about is ‘a computer program someone wrote.’”

[andy]

感覺reinforcement learning跟unsupervised learning都是讓機器無師自通, 請問他們之前的差別是什麼, 還是說reinforcement learning也算是一種unsupervised learning

[shadowue]

very awesome course, thanks for sharing

[李爺爺-yb]

在22分鐘0秒的時候，教授說的英文"Pixel" (像素)，被誤寫為"Peak Zone"。