Scribd
Upload
3 views

Lecture 02 - Supervised Learning

The document discusses supervised learning with a focus on predicting whether a car is a family car based on various features such as price and power. It outlines the model formulation, including regression techniques and error minimization strategies. The goal is to optimize the model parameters to improve prediction accuracy.

Uploaded by

boran.kaymakci
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Lecture 02 - Supervised Learning

The document discusses supervised learning with a focus on predicting whether a car is a family car based on various features such as price and power. It outlines the model formulation, including regression techniques and error minimization strategies. The goal is to optimize the model parameters to improve prediction accuracy.

Uploaded by

boran.kaymakci
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

supervised Learning

x =
&(xi ,
yi) training set .
Task : to predict whether a car is a
family cor ornot

00p00] B

I
210000
A
1

]
-

[xiz-engine
Xi1-price
.

Xi =
[1500-4000)power
xiEIR -
2x1

a family a
zi =
&xis similaritystance
e

min
I
Miff+diff power

xxy

TYz-sceling
X min
NXD
-

Tax-min
# of instancesHofdmesis a
mmmaxsclng
ensee ·: family cor

of

"
car
⑧
·:
other types
[0]
[]
22
y
=

x =

X22
x2 =
[2]32 = [1]

price i :
Model Family
RECTANGLES (p1
a =

pa , 21
223
FAMILY OF ,
,

W
perometers
model
LEARNING= the best R.
Finding Se -XN+, 2
p p e e fpXN +,
[
i
, , ,

3n + 1 = 1

f(xn )pped) = else


+ N+
YN + =
0
·
Regression
ideel solutio
E

NXX
x(mileage)
Model Fine Notwi R = Ew
ut

a
**
Wo
,
wh
*
!
y =
w + wix1 ·

32 = w .
wX21 + e2 =
32 -y2
XN +
: " i
M
wotwix es =
1

36-y0
&
. + W XN +,
=
36
N =
W
+1
X
X minimize
minimize

~minimize
en

ye 2

minimize -]
to :
with respect Wo we

& g(a)fall
,
2
a)
+

Ja
Error (wo ,
welx) = lyi-(utwexin)
-
awotwex)
Error

(gitw)
=
]
2 .

· wow
#I= (gi-(wotwexi)(xi)
* 2

-
-
.

-
-we
Exercis
d
a on
* -
Fin =
We

~ (N)-m[ -

You might also like