Baby Machine Learning

By Hes

Transcript

1. Hes 2021-06-23 Baby Machine Learning From Regression To Neural Network
2. Machine Learning •AI •Deep Learning •Classi fi cation •Regression •Clustering •Data Mining •Reinforcement Learning ...umm "MATH"
3. Machine Learning •AI •Deep Learning •Classi fi cation •Regression •Clustering •Data Mining •Reinforcement Learning ...umm "MATH"
4. How people think about House price • How many people? • How Big? • How many rooms & bathrooms? • Where the house is? + ठࣁӂ, ৉ࣁӂ
5. How people think about House price - Relation • How many people? • Size • # Rooms & Bathrooms • Where the house is? • How Big? • ठࣁӂ • ৉ࣁӂ • Family Size • Safety
6. How people think about House price - Priority • How many people? • Size • # Rooms & Bathrooms • Where the house is? • How Big? • ठࣁӂ • ৉ࣁӂ • Family Size • Safety 1
7. How people think about House price - Priority • How many people? • Size • # Rooms & Bathrooms • Where the house is? • How Big? • ठࣁӂ • ৉ࣁӂ • Family Size • Safety 1 Depends on a situation
8. Demand makes Price How people think about
9. Demand makes Price How people think about
10. makes Price Various Situation How people think about
11. Data Various Situation How people think about
12. Learn Method from Model variable Brain Data How people think about
13. Let's start with simple model
14. What makes price Overview Location point • ठࣁӂ • ৉ࣁӂ • Safety • Family Size • Size • Rooms &BR Volume point Family point Price
15. What makes price Overview Location point • ठࣁӂ • ৉ࣁӂ • Safety • Family Size • Size • Rooms &BR Volume point Family point Price
16. Linear Regression House price • ठࣁӂ • ৉ࣁӂ • Safety ? ? ? • Family Size ? Location Point
17. Linear Regression House price • ठࣁӂ • ৉ࣁӂ • Safety ? ? ? • Family Size ? Location Point Safetyо 95, ठࣁӂ੉ 1(True), ৉ࣁӂ 0(False), Family Sizeо 4ݺ੉ݶ, Location pointח 77੼੉ ա৬ঠ ೞפө... ੉ۧѱ ߄Լ઱ݶ!
18. Linear Regression House price • ठࣁӂ • ৉ࣁӂ • Safety ? ? ? • Family Size ? Location Point Ӓېࢲ ӒѦ ־о ೞחؘ?! ৘ஏ Ѿҗ৬ पઁ чਸ ઴ৈ઱ח ੌਸ.. Optimizerо ೞݶ غѷ׮!!
19. Linear Regression House price • ठࣁӂ • ৉ࣁӂ • Safety ? ? ? • Family Size ? Location Point Ӓېࢲ ӒѦ ־о ೞחؘ?! ৘ஏ Ѿҗ৬ पઁ чਸ ઴ৈ઱ח ੌਸ.. Optimizerо ೞݶ غѷ׮!!
20. Linear Regression Happy case
21. Linear Regression Umm.. case
22. Activation Fuction move on from Linear Linear Regression = Non-Linear Result + Activation(nonlinearities)
23. Activation Fuction move on from Linear Linear Regression = Non-Linear Result + Activation(nonlinearities)
24. σ( ) Location point Logistic Regression House price =
25. σ( ) Location point Logistic Regression House price • ठࣁӂ • ৉ࣁӂ • Safety ? ? ? • Family Size ? Good / Not Good (Location) = աח ݅઒ೡ ഛܫਸ ঌҊ रয.. Ӓېࢲ ഛܫ਷?
26. Neural Network House price - Stack the regression a( ) Location point • ठࣁӂ • ৉ࣁӂ • Safety • Family Size • Size • Rooms &BR a( ) Volume point a( ) Family point Price
27. Neural Network House price - Stack the regression a( ) Location point • ठࣁӂ • ৉ࣁӂ • Safety • Family Size • Size • Rooms &BR a( ) Volume point a( ) Family point Prob σ( )
28. Neural Network House price - Stack the regression a( ) Location point • ठࣁӂ • ৉ࣁӂ • Safety • Family Size • Size • Rooms &BR a( ) Volume point a( ) Family point Prob σ( ) Prob σ( ) Prob σ( ) ARGMAX
29. What we can do • Probability of Conversion • Classi fi cation • Scoring • Ranking …
30. ೞ૑݅ ৈ੹൤ য۵׮..