030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

MachinelearninginPython

Thankyou

Ireallyappreciatethehelp:)

Amazing!!

Whatthef**kisgoingon?!?

Thanksfornothing.Whatahorribleuserexperience

01.

Algorithms

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

2=>{2}

3=>{3}

4=>{2,2}

5=>{5}

6=>{2,3}

7=>{7}

8=>{2,2,2}

2=>{2}

3=>{3}

4=>{2,2}

5=>{5}

6=>{2,3}

7=>{7}

8=>{2,2,2}

2=>{2}

3=>{3}

4=>{2,2}

5=>{5}

6=>{2,3}

7=>{7}

8=>{2,2,2}

Q.E.D

02.MachineLearning

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

2=>{2}

3=>{7}

4=>{}

5=>{2,2,2,2,2,...}

6=>{“airplane”}

7=>{e,n}

8=>{?}

?

2=>{2}

3=>{7}

4=>{}

5=>{2,2,2,2,2,...}

6=>{“airplane”}

7=>{e,n}

8=>{?}

03.Linearregression

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

m2

NOK

72

4200000

68

3790000

72

3890000

42

3400000

66

4950000

55

3600000

f(x)=ax+b

f(x)=0x+3800000

f(x)=27272x+2145484

cost(x,f(x))={

0iff(x)=x1

cost(x,f(x))=(f(x)-x)2

04.Neuralnets

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

a

y

b

x

f(x)=ax+b

a

y

b

x

b

a1

a2

y

x1x2

b

a1

a2

∫

x1

y

x2

∫

∫

x1 y

∫

∫

∫

∫

x1 y

?

x1

x2 y

x3

x1

y1

x2

y2

x3

x1

Buttercup

x2

Daisy

x3

Buttercup

Daisy

Buttercup

Daisy

Buttercup:0.8

Daisy:0.2

Buttercup:0.8

Daisy:0.2

1.0

0.8

Buttercup:0.8

Daisy:0.2

1.0

0.8

05.ConvolutionalNeuralNets

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

ButtercupDaisy

}

}

}

}

}

ButtercupDaisy

Convolution

Pooling

Convolution

Convolution

Pooling

depth

height

width

ButtercupDaisy

06.Transferlearning

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

Tench

...

Goldfish

985.

Daisy

986.

Slipper

999.

Ear

...

1000.Toiletpaper

Tench

...

Goldfish

985.

Daisy

986.

Slipper

999.

Ear

...

1000.Toiletpaper

Bluebell

...

Buttercup

Fritilliary

...

Iris

Tulip

Windflower

07.Python

030918|MachinelearninginPython|

esten@epimed.ai

|epigrammedtech

1

0

5

0

3

4

9

0

12

4

6

19

2

1

27

4

2

0

8

3

8

5

11

1

13

8

4

6

7

3

1

0

2

12

0

4

y=ax+b

Servingimages

Preprocessing

Augmentations

Coremodel

Specificmodel

1 3 6 2 4/5

'bluebell'

0

[1,

0,

0,

0,

0,

0]

'buttercup'

1

[0,

1,

0,

0,

0,

0]

'