A Brief Introduction of Machine Learning
Published:
Perhaps it will be wise to approach the subject cautiously
A brief introduction of machine learning
Application of machine learning:
1.perform seb search
2.recognize human speech/speech recognition
3.diagnose diseases from x-rays
4.build a self-driving car
5.computer vision for street view images and advertising
6.augmented reality to combating payment forwards
…
——
What is machine learning:
give computers the ability to learn without being explicitly programmed
——
Main types of machine learning:
1.supervised learning(used most)
2.unsupervised learning
*Reinforcement Learning
——
Supervised learning:
learn from being given right answers
Input(X)→Output(Y)→Application
1.email→spam(yes or not)→spam filtering
2.audio→text transcripts→speech recognition
3.English→Chinese→machine translation
4.ad, user info→click(yes or not)→online advertising
5.image, radar info→position of other cars→self-driving car
6.image of phone→defect(yes or not)→visual inspection
…
give example of inputs X and right answers Y
then take a brand new input X something it never seen before and try to produce the appropriate corresponding output Y
—
Regression:
predict a number from all of the infinitely many number of possible numbers
e.g.
price Y and house size X
—
Classification:
predict categories, only two possible outputs(cat or dog, benign or malignant)
e.g.
figure out if a tumor that is a lump is malignant meaning cancerous or dangerous
—
Unsupervised learning:
find something interesting in unlabeled data
e.g.
not diagnore whether the tumor is benign or malignant but place unlabeled data into different clusters
1.online news(find the same words in different titles)
2.DNA micoarray(show the degree to which different individuals do or do not have a specific gene active)
3.grouping customers(group customers into different market segments)
…
formal definition:
data only comes with input X, but not output Y
algorithm has to find STRUCTURE in the data
—