1 Introduction

1.2 Welcome

Database mining
- Large datasets from growth of automation/web.
- E.g., Web click data, medical records, biology, engineering
Application cannot program by hand.
- E.g., Autonomous helicopter, handwriting recognition, most of Natural Language Processing (NLP), Computer Vision.
Self-customizing programs
- E.g., Amazon, Netflix product recommendations
Understanding human learning (brain, real AI).

Arthur Samuel (1959). Machine Learning: Field of study that gives computers the ability to learn without being explicitly programmed.
Tom Mitchell (1998) Well-posed Learning Problem: 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.

Others: Reinforcement learning, recommender systems.

Also talk about: Practice advice for applying learning algorigthms.

“right answers” given

Regression problem
- Predict continuous valued output (price)
- E.g., housing price prediction.
Classfication problem
- Discrete valued output (0 or 1, or more)
- Multiple/Infinite features might be considered, e.g., tumor size, age, clump thickness, uniformity of cell size, uniformity of cell shape
- E.g., breast cancer (malignant, benign),

no “right answers” given

E.g., Google News, Organize computing clusters, Social network analysis, Market segmentation, Astronomical data analysis

Cocktail party problem algorithm, in Octave:

[W,s,v] = svd((repmat(sum(x.*x,1),size(x,1),1).*x)*x');

Supervised Learning: Given the “right answer” for each example in the data.

Regression Problem: Predict real-valued output.

How to represent h ? \(h_θ(x) = Θ_0 + Θ_1x\)