【ML】1 Introduction
Lecture 1 Introduction
What is machine learning
Definition
Arthur Samuel(1959): Field of study that gives computers the ability to learn without being explicitly programmed.
-
Tom Mitchell (1998)
-
A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E.
- E:经验,重复训练中获得的数据
- T:任务,所重复的训练
- P:性能度量,符合预期结果的概率
Machine learning algorithms:
- Supervised learning:人工设定计算机怎么学习
- Unsupervised learning:计算机自己学习
- Reinforcement learning
- recommender systems
Supervised Learning
Example
-
预测房价
获取数据集,房价及房子面积,获得拟合数据,得以预测——回归问题(Regression)
-
预测肿瘤良性或恶性
获取数据集,肿瘤大小及其为良性或是恶性,进行分类(一组有限离散数据),计算某大小的肿瘤为良性的概率或为恶性的概率——分类问题(Classification)
可能以多种属性进行分类,本例可能有肿瘤性质(需要判断的问题)、患者年龄、肿瘤大小、肿瘤形状
Definition
- "right answers" given(给出标签)
- 常见问题类型:回归问题(Regression)、分类问题(Classification)
Unsupervised Learning
Example
-
聚类算法
同类新闻分类、检测特定基因表达程度、大型计算机集群、社交网络分析、市场细分
-
鸡尾酒会问题
多个麦克风收集鸡尾酒会音频(声音很混乱),辨别音源(每个人说的话)
-
Cocktail party problem algorithm
[W,s,v] = svd((repmat(sum(x.*x,1),size(x,1),1).*x)*x');
-

浙公网安备 33010602011771号