MACHINE LEARNING - FOUNDATIONS

TUTORIAL - WEEK 2

IIT Madras Online Degree

Outline

1. LINEAR APPROXIMATION

2. HIGHER ORDER APPROXIMATIONS

3. MULTIVARIATE LINEAR APPROXIMATION

4. DIRECTIONAL DERIVATIVES

2
LINEAR APPROXIMATION
Linear approximation (Linearization)

Def:
Approximation of any function using a linear function .

4
Linear approximation (Linearization)

Def:
Approximation of any function using a linear function .

Need:
• Linear functions are easier to work with.

4
Linear approximation (Linearization)

Def:
Approximation of any function using a linear function .

Need:
• Linear functions are easier to work with.
• Finding approximate values of functions at certain points when exact values are not
known. 4
The equation

The linear approximation 𝐿(𝑥) of a function 𝑓(𝑥) at point 𝑎 is given by:

′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

5
The equation

The linear approximation 𝐿(𝑥) of a function 𝑓(𝑥) at point 𝑎 is given by:

′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

This is indeed the equation of a tangent line:

𝑦 − 𝑦1 = 𝑚(𝑥 − 𝑥1 )
𝑦 = 𝑦1 + 𝑚(𝑥 − 𝑥1 )

5
The equation

The linear approximation 𝐿(𝑥) of a function 𝑓(𝑥) at point 𝑎 is given by:

′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

This is indeed the equation of a tangent line:

𝑦 − 𝑦1 = 𝑚(𝑥 − 𝑥1 )
𝑦 = 𝑦1 + 𝑚(𝑥 − 𝑥1 )

If 𝑥1 = 𝑎, 𝑦1 = 𝑓(𝑎) and 𝑚 = 𝑓 ′ (𝑎), we get,

𝑦 = 𝑓(𝑎) + 𝑓 ′ (𝑎)(𝑥 − 𝑎)

5
Problem 1
√
Compute the approximate value of 50.

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.

′ 1 −1
𝑓 (𝑥) = 𝑥 2
2

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.

′ 1 −1
𝑓 (𝑥) = 𝑥 2
2
√
𝑓(49) = 49 = 7

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.

′ 1 −1
𝑓 (𝑥) = 𝑥 2
2
√
𝑓(49) = 49 = 7
′ 1 1
𝑓 (49) = √ =
(2)( 49) 14

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.

′ 1 −1
𝑓 (𝑥) = 𝑥 2
2
√
𝑓(49) = 49 = 7
′ 1 1
𝑓 (49) = √ =
(2)( 49) 14
′
𝐿(𝑥) = 𝑓(49) + 𝑓 (49)(𝑥 − 49)

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√ 10
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.
8
′ 1 −1 𝑎 = 49
𝑓 (𝑥) = 𝑥 2
2
√ 6
𝑓(49) = 49 = 7
′ 1 1 4
𝑓 (49) = √ =
(2)( 49) 14 √
2 𝑓(𝑥) = 𝑥
′
𝐿(𝑥) = 𝑓(49) + 𝑓 (49)(𝑥 − 49) 𝐿(𝑥) = 7 + 𝑥/14 − 49/14
1 20 40 60 80 100
𝐿(𝑥) = 7 + (𝑥 − 49)
14

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√ 10
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.
8
′ 1 −1 𝑎 = 49
𝑓 (𝑥) = 𝑥 2
2
√ 6
𝑓(49) = 49 = 7
′ 1 1 4
𝑓 (49) = √ =
(2)( 49) 14 √
2 𝑓(𝑥) = 𝑥
′
𝐿(𝑥) = 𝑓(49) + 𝑓 (49)(𝑥 − 49) 𝐿(𝑥) = 7 + 𝑥/14 − 49/14
1 20 40 60 80 100
𝐿(𝑥) = 7 + (𝑥 − 49)
14
√ 1 1
Approximate value of 50 = 𝐿(50) = 7 + 14 (50 − 49) = 7 + 14 = 7.071

6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√ 10
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.
8
′ 1 −1 𝑎 = 49
𝑓 (𝑥) = 𝑥 2
2
√ 6
𝑓(49) = 49 = 7
′ 1 1 4
𝑓 (49) = √ =
(2)( 49) 14 √
2 𝑓(𝑥) = 𝑥
′
𝐿(𝑥) = 𝑓(49) + 𝑓 (49)(𝑥 − 49) 𝐿(𝑥) = 7 + 𝑥/14 − 49/14
1 20 40 60 80 100
𝐿(𝑥) = 7 + (𝑥 − 49)
14
√ 1 1
Approximate value of 50 = 𝐿(50) = 7 + 14 (50 − 49) = 7 + 14 = 7.071
√
• Note 1: Actual value of 50 (up to 3 decimal places) is 7.071. 6
Problem 1
√
Compute the approximate value of 50.
√ √
The closest known value to 50 is 49, so
√ 10
we set 𝑓(𝑥) = 𝑥 and 𝑎 = 49.
8
′ 1 −1 𝑎 = 49
𝑓 (𝑥) = 𝑥 2
2
√ 6
𝑓(49) = 49 = 7
′ 1 1 4
𝑓 (49) = √ =
(2)( 49) 14 √
2 𝑓(𝑥) = 𝑥
′
𝐿(𝑥) = 𝑓(49) + 𝑓 (49)(𝑥 − 49) 𝐿(𝑥) = 7 + 𝑥/14 − 49/14
1 20 40 60 80 100
𝐿(𝑥) = 7 + (𝑥 − 49)
14
√ 1 1
Approximate value of 50 = 𝐿(50) = 7 + 14 (50 − 49) = 7 + 14 = 7.071
√
• Note 1: Actual value of 50 (up to 3 decimal places) is 7.071.
√ 6
• Note 2: 𝐿(100) gives 10.64 while the actual value of 100 is 10.
Problem 2

Compute the approximate value of 𝑒0.017 .

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
′
𝑓 (𝑥) = 𝑒𝑥

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′
𝑓 (0) = 1

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′
𝑓 (0) = 1
′
𝐿(𝑥) = 𝑓(0) + 𝑓 (0)(𝑥 − 0)

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we 𝑓(𝑥) = 𝑒𝑥
𝐿(𝑥) = 1 + 𝑥
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
100
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′ 50
𝑓 (0) = 1
′
𝐿(𝑥) = 𝑓(0) + 𝑓 (0)(𝑥 − 0)
𝑎=0
𝐿(𝑥) = 1 + 1(𝑥) = 1 + 𝑥 −4 −2 2 4

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we 𝑓(𝑥) = 𝑒𝑥
𝐿(𝑥) = 1 + 𝑥
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
100
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′ 50
𝑓 (0) = 1
′
𝐿(𝑥) = 𝑓(0) + 𝑓 (0)(𝑥 − 0)
𝑎=0
𝐿(𝑥) = 1 + 1(𝑥) = 1 + 𝑥 −4 −2 2 4

Approximate value of 𝑒 0.017

= 𝐿(0.017) = 1 + 0.017 = 1.017

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we 𝑓(𝑥) = 𝑒𝑥
𝐿(𝑥) = 1 + 𝑥
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
100
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′ 50
𝑓 (0) = 1
′
𝐿(𝑥) = 𝑓(0) + 𝑓 (0)(𝑥 − 0)
𝑎=0
𝐿(𝑥) = 1 + 1(𝑥) = 1 + 𝑥 −4 −2 2 4

Approximate value of 𝑒 0.017

= 𝐿(0.017) = 1 + 0.017 = 1.017
• Note 1: Actual value of 𝑒0.017 is also 1.017.

7
Problem 2

Compute the approximate value of 𝑒0.017 .

The closest known value to 𝑒0.017 is 𝑒0 , so we 𝑓(𝑥) = 𝑒𝑥
𝐿(𝑥) = 1 + 𝑥
set 𝑓(𝑥) = 𝑒𝑥 and 𝑎 = 0.
100
′
𝑓 (𝑥) = 𝑒𝑥
𝑓(0) = 1
′ 50
𝑓 (0) = 1
′
𝐿(𝑥) = 𝑓(0) + 𝑓 (0)(𝑥 − 0)
𝑎=0
𝐿(𝑥) = 1 + 1(𝑥) = 1 + 𝑥 −4 −2 2 4

Approximate value of 𝑒 0.017

= 𝐿(0.017) = 1 + 0.017 = 1.017
• Note 1: Actual value of 𝑒0.017 is also 1.017.
• Note 2: 𝐿(1) gives 2 while the actual value of 𝑒 is 2.718.
7
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
𝑎 = 5.
′ 1 1
𝑓 (𝑥) = (𝑥 + 4)− 2
2

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
𝑎 = 5.
′ 1 1
𝑓 (𝑥) = (𝑥 + 4)− 2
2
√
𝑓(5) = 5+4=3

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
𝑎 = 5.
′ 1 1
𝑓 (𝑥) = (𝑥 + 4)− 2
2
√
𝑓(5) = 5+4=3
′ 1 1
𝑓 (5) = √ =
(2)( 9) 6

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
𝑎 = 5.
′ 1 1
𝑓 (𝑥) = (𝑥 + 4)− 2
2
√
𝑓(5) = 5+4=3
′ 1 1
𝑓 (5) = √ =
(2)( 9) 6
′
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5)

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
4
𝑎 = 5.
′ 1 1 𝑎=5
𝑓 (𝑥) = (𝑥 + 4)− 2 3
2
√
𝑓(5) = 5+4=3 2
′ 1 1
𝑓 (5) = √ = √
(2)( 9) 6 1
𝑓(𝑥) = 𝑥 + 4
′ 𝐿(𝑥) = 3 + 𝑥/6 − 5/6
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5)
1 𝑥−5 −5 5 10 15
𝐿(𝑥) = 3 + (𝑥 − 5) = 3 +
6 6

8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
4
𝑎 = 5.
′ 1 1 𝑎=5
𝑓 (𝑥) = (𝑥 + 4)− 2 3
2
√
𝑓(5) = 5+4=3 2
′ 1 1
𝑓 (5) = √ = √
(2)( 9) 6 1
𝑓(𝑥) = 𝑥 + 4
′ 𝐿(𝑥) = 3 + 𝑥/6 − 5/6
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5)
1 𝑥−5 −5 5 10 15
𝐿(𝑥) = 3 + (𝑥 − 5) = 3 +
6 6
6−5 1 19
Approximate value of 𝑓(6) = 𝐿(6) = 3 + 6 =3+ 6 = 6 = 3.1666
8
Problem 3

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √
The closest known value to 𝑥 + 4 is 9
√
with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and
4
𝑎 = 5.
′ 1 1 𝑎=5
𝑓 (𝑥) = (𝑥 + 4)− 2 3
2
√
𝑓(5) = 5+4=3 2
′ 1 1
𝑓 (5) = √ = √
(2)( 9) 6 1
𝑓(𝑥) = 𝑥 + 4
′ 𝐿(𝑥) = 3 + 𝑥/6 − 5/6
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5)
1 𝑥−5 −5 5 10 15
𝐿(𝑥) = 3 + (𝑥 − 5) = 3 +
6 6
Approximate value of 𝑓(6) = 𝐿(6) = 3 + 6−5
6 =3+ 6 =
1 19
6 = 3.1666
√ 8
• Note: Actual value of 𝑓(6) = 10 is 3.1622. Why?
HIGHER ORDER APPROXIMATIONS
Higher order approximations

Linear Approximation
′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

10
Higher order approximations

Linear Approximation
′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

Quadratic Approximation
″
′ 𝑓 (𝑎)
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎) + (𝑥 − 𝑎)2
2

10
Higher order approximations

Linear Approximation
′
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎)

Quadratic Approximation
″
′ 𝑓 (𝑎)
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (𝑎)(𝑥 − 𝑎) + (𝑥 − 𝑎)2
2

Higher-order Approximations
𝑓 (2) (𝑎)
𝐿(𝑥) = 𝑓(𝑎) + 𝑓 (1) (𝑎)(𝑥 − 𝑎) + (𝑥 − 𝑎)2 +
2
𝑓 (3) (𝑎) 𝑓 (4) (𝑎)
+ (𝑥 − 𝑎)3 + (𝑥 − 𝑎)4 ...
3⋅2 4⋅3⋅2

10
Problem 4

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?

11
Problem 4

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √ √
The closest known value to 𝑥 + 4 is 9 with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and 𝑎 = 5.

11
Problem 4

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √ √
The closest known value to 𝑥 + 4 is 9 with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and 𝑎 = 5.
′ 1 1
𝑓 (𝑥) =(𝑥 + 4)− 2
2
″ 1 3
𝑓 (𝑥) = − (𝑥 + 4)− 2
4
‴ 3 5
𝑓 (𝑥) = (𝑥 + 4)− 2
8

11
Problem 4

√
Let 𝑓(𝑥) = 𝑥 + 4, what is 𝑓(6)?
√ √ √
The closest known value to 𝑥 + 4 is 9 with 𝑥 = 5, so we set 𝑓(𝑥) = 𝑥 + 4 and 𝑎 = 5.
′ 1 1
𝑓 (𝑥) = (𝑥 + 4)− 2
2
″ 1 3
𝑓 (𝑥) = − (𝑥 + 4)− 2
4
‴ 3 5
𝑓 (𝑥) = (𝑥 + 4)− 2
8
√
𝑓(5) = 5+4=3
′ 1 1
𝑓 (5) = √ =
(2)( 9) 6
″ 1
𝑓 (5) = −
108
‴ 1
𝑓 (5) = 11
(24)(27)
 ″ ‴
′ 𝑓 (5) 𝑓 (5)
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5) + (𝑥 − 5)2 + (𝑥 − 5)3 + ...
2 (3)(2)
1 1 1
𝐿(𝑥) = 3 + (𝑥 − 5) − (𝑥 − 5)2 + (𝑥 − 5)3
6 (108)(2) (24)(27)(6)

12
 ″ ‴
′ 𝑓 (5) 𝑓 (5)
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5) + (𝑥 − 5)2 + (𝑥 − 5)3 + ...
2 (3)(2)
1 1 1
𝐿(𝑥) = 3 + (𝑥 − 5) − (𝑥 − 5)2 + (𝑥 − 5)3
6 (108)(2) (24)(27)(6)

𝑎=5
3

√
1 𝑓(𝑥) = 𝑥 + 4
𝐿(𝑥)

−5 5 10 15 20

Approximate value of
𝑓(6) = 𝐿(6) = 3 + 61 (6 − 5) − 1
(108)(2) (6 − 5)2 + 1
(24)(27)(6) (6 − 5)3 = 3.1622 12
 ″ ‴
′ 𝑓 (5) 𝑓 (5)
𝐿(𝑥) = 𝑓(5) + 𝑓 (5)(𝑥 − 5) + (𝑥 − 5)2 + (𝑥 − 5)3 + ...
2 (3)(2)
1 1 1
𝐿(𝑥) = 3 + (𝑥 − 5) − (𝑥 − 5)2 + (𝑥 − 5)3
6 (108)(2) (24)(27)(6)

𝑎=5
3

√
1 𝑓(𝑥) = 𝑥 + 4
𝐿(𝑥)

−5 5 10 15 20

Approximate value of
𝑓(6) = 𝐿(6) = 3 + 61 (6 − 5) − 1
(108)(2) (6 − 5)2 + 1
(24)(27)(6) (6 − 5)3 = 3.1622 12
√
MULTIVARIATE LINEAR APPROXIMA-
TION
Linear approximation of functions involving multiple variables

The linear approximation of a function 𝑓 of two variables 𝑥 and 𝑦 in the neighborhood of

(𝑎, 𝑏) is:
𝜕𝑓 𝜕𝑓
𝐿(𝑥, 𝑦) = 𝑓(𝑎, 𝑏) + (𝑎, 𝑏)(𝑥 − 𝑎) + (𝑎, 𝑏)(𝑦 − 𝑏)
𝜕𝑥 𝜕𝑦

14
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

15
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑦 + 𝑒𝑥𝑦 = 𝑥𝑦𝑒𝑥𝑦 + 𝑒𝑥𝑦
𝜕𝑥
𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑥 = 𝑥2 𝑒𝑥𝑦
𝜕𝑦

15
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑦 + 𝑒𝑥𝑦 = 𝑥𝑦𝑒𝑥𝑦 + 𝑒𝑥𝑦
𝜕𝑥
𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑥 = 𝑥2 𝑒𝑥𝑦
𝜕𝑦
Here (𝑎, 𝑏) = (1, 0).

15
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑦 + 𝑒𝑥𝑦 = 𝑥𝑦𝑒𝑥𝑦 + 𝑒𝑥𝑦
𝜕𝑥
𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑥 = 𝑥2 𝑒𝑥𝑦
𝜕𝑦
Here (𝑎, 𝑏) = (1, 0).
𝑓(1, 0) = 𝑒0 = 1

15
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑦 + 𝑒𝑥𝑦 = 𝑥𝑦𝑒𝑥𝑦 + 𝑒𝑥𝑦
𝜕𝑥
𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑥 = 𝑥2 𝑒𝑥𝑦
𝜕𝑦
Here (𝑎, 𝑏) = (1, 0).
𝑓(1, 0) = 𝑒0 = 1
𝜕𝑓 𝜕𝑓
(𝑎, 𝑏) = (1, 0) = 𝑒0 = 1
𝜕𝑥 𝜕𝑥

15
Problem 5

Find the linearization of 𝑓(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 at (1, 0). Use it to approximate 𝑓(1.1, −0.1).

𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑦 + 𝑒𝑥𝑦 = 𝑥𝑦𝑒𝑥𝑦 + 𝑒𝑥𝑦
𝜕𝑥
𝜕𝑓
(𝑥, 𝑦) = 𝑥𝑒𝑥𝑦 𝑥 = 𝑥2 𝑒𝑥𝑦
𝜕𝑦
Here (𝑎, 𝑏) = (1, 0).
𝑓(1, 0) = 𝑒0 = 1
𝜕𝑓 𝜕𝑓
(𝑎, 𝑏) = (1, 0) = 𝑒0 = 1
𝜕𝑥 𝜕𝑥
𝜕𝑓 𝜕𝑓
(𝑎, 𝑏) = (1, 0) = 𝑒0 = 1
𝜕𝑦 𝜕𝑦

15
 𝜕𝑓 𝜕𝑓
𝐿(𝑥, 𝑦) = 𝑓(1, 0) + (1, 0)(𝑥 − 1) + (1, 0)(𝑦 − 0)
𝜕𝑥 𝜕𝑦

16
 𝜕𝑓 𝜕𝑓
𝐿(𝑥, 𝑦) = 𝑓(1, 0) + (1, 0)(𝑥 − 1) + (1, 0)(𝑦 − 0)
𝜕𝑥 𝜕𝑦
= 1 + 1(𝑥 − 1) + 1(𝑦)
= 𝑥+𝑦

16
 𝜕𝑓 𝜕𝑓
𝐿(𝑥, 𝑦) = 𝑓(1, 0) + (1, 0)(𝑥 − 1) + (1, 0)(𝑦 − 0)
𝜕𝑥 𝜕𝑦
= 1 + 1(𝑥 − 1) + 1(𝑦)
= 𝑥+𝑦
𝑓(1.1, −0.1) = 𝐿(1.1, −0.1)
= 1.1 − 0.1 = 1

16
 𝜕𝑓 𝜕𝑓
𝐿(𝑥, 𝑦) = 𝑓(1, 0) + (1, 0)(𝑥 − 1) + (1, 0)(𝑦 − 0)
𝜕𝑥 𝜕𝑦
= 1 + 1(𝑥 − 1) + 1(𝑦)
= 𝑥+𝑦
𝑓(1.1, −0.1) = 𝐿(1.1, −0.1)
= 1.1 − 0.1 = 1
1.1
The actual value of f(1.1, -0.1) = 1.1e−0.11 = 1.11628 = 0.98542

16
DIRECTIONAL DERIVATIVES
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦)

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦)

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑦 (keeping 𝑥 fixed).

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑦 (keeping 𝑥 fixed).
• Directional derivative of 𝑓(𝑥, 𝑦)

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑦 (keeping 𝑥 fixed).
• Directional derivative of 𝑓(𝑥, 𝑦) = Rate of change of 𝑓 if we allow both 𝑥 and 𝑦 to
change simultaneously (in some direction (𝑢)).

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑦 (keeping 𝑥 fixed).
• Directional derivative of 𝑓(𝑥, 𝑦) = Rate of change of 𝑓 if we allow both 𝑥 and 𝑦 to
change simultaneously (in some direction (𝑢)).

𝐷𝑢⃗⃗⃗ ⃗ 𝑓(𝑥, 𝑦) = ∇𝑓 ⋅ 𝑢
𝜕𝑓 𝜕𝑓
=[ , ] ⋅ [𝑢1 , 𝑢2 ]
𝜕𝑥 𝜕𝑦
𝜕𝑓 𝜕𝑓
= 𝑢1 + 𝑢2
𝜕𝑥 𝜕𝑦

18
Directional Derivative

𝜕𝑓
• 𝑓𝑥 (𝑥, 𝑦) = 𝜕𝑥 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑥 (keeping 𝑦 fixed).
𝜕𝑓
• 𝑓𝑦 (𝑥, 𝑦) = 𝜕𝑦 (𝑥, 𝑦) = Rate of change of 𝑓 as we vary 𝑦 (keeping 𝑥 fixed).
• Directional derivative of 𝑓(𝑥, 𝑦) = Rate of change of 𝑓 if we allow both 𝑥 and 𝑦 to
change simultaneously (in some direction (𝑢)).

𝐷𝑢⃗⃗⃗ ⃗ 𝑓(𝑥, 𝑦) = ∇𝑓 ⋅ 𝑢
𝜕𝑓 𝜕𝑓
=[ , ] ⋅ [𝑢1 , 𝑢2 ]
𝜕𝑥 𝜕𝑦
𝜕𝑓 𝜕𝑓
= 𝑢1 + 𝑢2
𝜕𝑥 𝜕𝑦

Directional derivative can be considered to be a weighted sum of partial derivatives.

18
Problem 6

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥𝑐𝑜𝑠(𝑦) in the direction of 𝑢

⃗⃗⃗⃗ = [2, 1].

19
Problem 6

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥𝑐𝑜𝑠(𝑦) in the direction of 𝑢

⃗⃗⃗⃗ = [2, 1].

𝜕𝑓
= cos(𝑦)
𝜕𝑥
𝜕𝑓
= −𝑥 sin(𝑦)
𝜕𝑦

19
Problem 6

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥𝑐𝑜𝑠(𝑦) in the direction of 𝑢

⃗⃗⃗⃗ = [2, 1].

𝜕𝑓
= cos(𝑦)
𝜕𝑥
𝜕𝑓
= −𝑥 sin(𝑦)
𝜕𝑦

⃗⃗⃗⃗ = [ √25 , √15 ]

Unit vector in the direction of 𝑢

19
Problem 6

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥𝑐𝑜𝑠(𝑦) in the direction of 𝑢

⃗⃗⃗⃗ = [2, 1].

𝜕𝑓
= cos(𝑦)
𝜕𝑥
𝜕𝑓
= −𝑥 sin(𝑦)
𝜕𝑦

⃗⃗⃗⃗ = [ √25 , √15 ]

Unit vector in the direction of 𝑢

𝜕𝑓 𝜕𝑓
𝐷𝑢⃗⃗⃗ ⃗ 𝑓(𝑥, 𝑦) = 𝑢1 + 𝑢2
𝜕𝑥 𝜕𝑦
2 1
= √ cos(𝑦) − √ 𝑥 sin(𝑦)
5 5

19
Problem 7

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥2 − 𝑥𝑦 in the direction of 𝑢

⃗⃗⃗⃗ = 0.6𝑖 + 0.8𝑗 at the point
(2, −3).

20
Problem 7

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥2 − 𝑥𝑦 in the direction of 𝑢

⃗⃗⃗⃗ = 0.6𝑖 + 0.8𝑗 at the point
(2, −3).
𝜕𝑓
= 2𝑥 − 𝑦
𝜕𝑥
𝜕𝑓
= −𝑥
𝜕𝑦

20
Problem 7

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥2 − 𝑥𝑦 in the direction of 𝑢

⃗⃗⃗⃗ = 0.6𝑖 + 0.8𝑗 at the point
(2, −3).
𝜕𝑓
= 2𝑥 − 𝑦
𝜕𝑥
𝜕𝑓
= −𝑥
𝜕𝑦
⃗⃗⃗⃗ is already a unit vector. 𝑢1 = 0.6, 𝑢2 = 0.8.
𝑢

20
Problem 7

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥2 − 𝑥𝑦 in the direction of 𝑢

⃗⃗⃗⃗ = 0.6𝑖 + 0.8𝑗 at the point
(2, −3).
𝜕𝑓
= 2𝑥 − 𝑦
𝜕𝑥
𝜕𝑓
= −𝑥
𝜕𝑦
⃗⃗⃗⃗ is already a unit vector. 𝑢1 = 0.6, 𝑢2 = 0.8.
𝑢
𝐷𝑢⃗⃗⃗ ⃗ 𝑓(𝑥, 𝑦) = 0.6(2𝑥 − 𝑦) + 0.8(−𝑥)
= 0.6(2𝑥 − 𝑦) − 0.8𝑥

20
Problem 7

Find the derivative of 𝑓(𝑥, 𝑦) = 𝑥2 − 𝑥𝑦 in the direction of 𝑢

⃗⃗⃗⃗ = 0.6𝑖 + 0.8𝑗 at the point
(2, −3).
𝜕𝑓
= 2𝑥 − 𝑦
𝜕𝑥
𝜕𝑓
= −𝑥
𝜕𝑦
⃗⃗⃗⃗ is already a unit vector. 𝑢1 = 0.6, 𝑢2 = 0.8.
𝑢
𝐷𝑢⃗⃗⃗ ⃗ 𝑓(𝑥, 𝑦) = 0.6(2𝑥 − 𝑦) + 0.8(−𝑥)
= 0.6(2𝑥 − 𝑦) − 0.8𝑥

𝐷𝑢⃗⃗⃗ ⃗ 𝑓(2, −3) = 0.6(2(2) + 3) − 0.8(2)

= 0.6(7) − 1.6
= 4.2 − 1.6
20
= 2.6
Thank you.