Artificial Intelligence

• A computer system solving real world problems by mimicking human intelligence

Machine Learning

• One of the AI methods learns pattern from sampled data

Deep Learning

• One of the ML methods based on artificial neutral network

Why Machine Learning

• We are living in the Big Data era

• There are lots of data available to train machine learning models

• CPU computing (중앙처리장치)
  - We can easily use the massive amounts of computing resources with cloud computing

• GPU Computing enhances the computing performances

• Business value creation with AI, Machine Learning
  - Google, Facebook, Netflix ...

Definition of Machine Learning

• Machine Learning is the study of computer algorithms that allow computer programs th automatically improve through experience

Composition of learning systems

• Environment : learning systems interacts with environment to accumulate experience
• Data : the memorized experiences interacting with the enviroments
• Model : a function f(x) represents the pattern of data
• Performance : evaluation criteria for the learning system. The system optimizes the performances to solve the problems

Model Evalution and Performance

• We must train the function for appropriately demonstrating the relationship between input and output variables
• MSE : Mean Squared Error
  - (실제값 - 예측값) 제곱의 평균
  - 오차가 작은 것이 더 좋은 모형

reference : K-MOOC 실습으로 배우는 머신러닝