Quantum Machine Learning

Amira Abbas
Quantum Research Group
University of KwaZulu-Natal
 Quantum Research Group

● Quantum computing, key distribution (teleportation),

open systems, evolutionary algorithms …


● Global/local partnerships


● Programmes/conferences/summer schools 


● Large focus on QML

Amira Abbas Wits QC SS 2019

What is QML and why should we care about it?
 Quantum machine learning

Amira Abbas Source: Supervised Learning on Quantum Computers: Schuld & Petruccione 2018 Wits QC SS 2019
 Quantum-classical machine learning

Amira Abbas Wits QC SS 2019

 Quantum-classical machine learning

● Can QCs offer something different?


● What “hard” stuff can we outsource to quantum computers?


- Hybrid computation


Amira Abbas Source: pennylane.ai Wits QC SS 2019

 Today

● Go through a small example of quantum classification


● Revisit some classical ML techniques


● Obtain a high level overview of variational circuits


● Examine a quantum algorithm that we can actually implement

which has real applications


Amira Abbas Wits QC SS 2019

 Squared-distance classifier

● Titanic data set (kaggle):


- have information on ticket price, cabin number and survival


● Problem: 

- want to predict if a passenger survived or not given how much they paid
and their cabin number


Amira Abbas Source: Supervised Learning on Quantum Computers: Schuld & Petruccione 2018 Wits QC SS 2019
 Squared-distance classifier

[0910] [2105]
8500 1200
x1 = y1 = 1 x2 = y2 = 0

Amira Abbas Wits QC SS 2019

 Squared-distance classifier

[0910] [2105]
8500 1200
x1 = y1 = 1 x2 = y2 = 0

[0.390] [0.990]
0.921 0.141
x1 = y1 = 1 x2 = y2 = 0

Amira Abbas Wits QC SS 2019

 Squared-distance classifier

[0910] [2105]
8500 1200
x1 = y1 = 1 x2 = y2 = 0

[0.390] [0.990]
0.921 0.141
x1 = y1 = 1 x2 = y2 = 0

[0.500]
0.866 y new = ?
x new =

Amira Abbas Wits QC SS 2019

 Squared-distance classifier

Amira Abbas Wits QC SS 2019

 Squared-distance classifier

1 new
wm = 1 − | x − x m |2 for M data points
c

Amira Abbas Wits QC SS 2019

 Squared-distance classifier

1 new
wm = 1 − | x − x m |2 for M data points
c

11
pxnew(y new = 0) =
∑
wm
χ i m|y m
=0

1 1
pxnew(y new = 1) =
∑
wm
χ M − i m|y m=1

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● We can use interference with a Hadamard transform to create distance

measures

Amira Abbas Source: https://arxiv.org/abs/1703.10793 Wits QC SS 2019

 Quantum squared-distance classifier

● How do we encode data into a quantum computer?

[0.390] [0.990]
0.921 0.141
x1 = y 1
=1 x2 = y 2
=0

[0.500] [0.500]
0.866 0.866
x new = x new =

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● Amplitude encoding

[0.390] [0.500]
0.921 0.866
x1 = x new =

[0.990] [0.500]
0.141 0.866
x2 = x new =

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● Amplitude encoding
1 | 000⟩

[0.390] [0.500]
0.921 0.866 1 | 001⟩
x1 = x new =
1 | 010⟩
1 1 | 011⟩

[0.990] [0.500]
x2 =
0.141
x new =
0.866 8 1 | 100⟩
1 | 101⟩
1 | 110⟩
1 | 111⟩

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● Amplitude encoding
0.921 | 000⟩

[0.390] [0.500]
0.921 0.866 0.390 | 001⟩
x1 = x new = | 010⟩
0.141
0.990 | 011⟩

[0.990] [0.500]
0.141 0.866 0.866 | 100⟩
x2 = x new =
0.500 | 101⟩
0.866 | 110⟩
0.500 | 111⟩

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● Amplitude encoding
0.921 | 000⟩ y1 = 1

[0.390] [0.500]
0.921 0.866 0.390 | 001⟩ y1 = 1
x1 = x new = | 010⟩ y2 = 0
0.141
0.990 | 011⟩ y2 = 0

[0.990] [0.500]
0.141 0.866 0.866 | 100⟩
x2 = x new =
0.500 | 101⟩
0.866 | 110⟩
0.500 | 111⟩

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

● Amplitude encoding
0.921 | 000⟩ y1 = 1

[0.390] [0.500]
0.921 0.866 0.390 | 001⟩ y1 = 1
x1 = x new = | 010⟩ y2 = 0
0.141
0.990 | 011⟩ y2 = 0

[0.990] [0.500]
0.141 0.866 0.866 | 100⟩ y1 = 1
x2 = x new = y1 = 1
0.500 | 101⟩
0.866 | 110⟩ y2 = 0
0.500 | 111⟩ y2 = 0

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

Index Label
0.921 | 000⟩ | 1⟩
0.390 | 001⟩ | 1⟩
0.141 | 010⟩ | 0⟩
0.990 | 011⟩ | 0⟩
0.866 | 100⟩ | 1⟩
0.500 | 101⟩ | 1⟩
0.866 | 110⟩ | 0⟩
0.500 | 111⟩ | 0⟩

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

11 1
px new(y new = 0) = (1 − | x new − x m |2 )
∑
χ i m|y c
m =0 The measurement statistics of
the label qubit are the same
as the result of the classical
squared distance classifier!

1 1 1
px new(y new = 1) = (1 − | x new − x m |2 )
∑
χ M − i m|y m =1 c

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

index1 | 0⟩
index2 | 0⟩
index3 | 0⟩
label | 0⟩

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

index1 | 0⟩ H

index2 | 0⟩ H

index3 | 0⟩ H

label | 0⟩
| 000⟩
| 001⟩ 1
1
| 010⟩
1
| 011⟩ 1 1
→
| 100⟩ 8 1
| 101⟩ 1
1
| 110⟩ 1
Amira Abbas | 111⟩ Wits QC SS 2019
 Quantum squared-distance classifier

index1 | 0⟩ H RY

index2 | 0⟩ H RY

index3 | 0⟩ H
RY
label | 0⟩
0.921 | 000⟩
Encode data
0.390 | 001⟩
0.141 | 010⟩
0.990 | 011⟩
0.866 | 100⟩
0.500 | 101⟩
0.866 | 110⟩
0.500 | 111⟩
Amira Abbas Wits QC SS 2019
 Quantum squared-distance classifier

index1 | 0⟩ H RY

index2 | 0⟩ H RY

index3 | 0⟩ H
RY
label | 0⟩ X X X X

Encode labels 0.921 | 000⟩ | 1⟩

0.390 | 001⟩ | 1⟩
0.141 | 010⟩ | 0⟩
0.990 | 011⟩ | 0⟩
0.866 | 100⟩ | 1⟩
0.500 | 101⟩ | 1⟩
0.866 | 110⟩ | 0⟩
Amira Abbas 0.500 | 111⟩ | 0⟩ Wits QC SS 2019
 Quantum squared-distance classifier

index1 | 0⟩ H RY H

index2 | 0⟩ H RY

index3 | 0⟩ H
RY
label | 0⟩ X X X X

0.921 | 000⟩ | 1⟩
Hadamard
0.390 | 001⟩ | 1⟩
transform
0.141 | 010⟩ | 0⟩
0.990 | 011⟩ | 0⟩
0.866 | 100⟩ | 1⟩
0.500 | 101⟩ | 1⟩
0.866 | 110⟩ | 0⟩
0.500 | 111⟩ | 0⟩
Amira Abbas Wits QC SS 2019
 Quantum squared-distance classifier

index1 | 0⟩ H RY H

index2 | 0⟩ H RY

index3 | 0⟩ H
RY
label | 0⟩ X X X X

Conditional
measurement

Amira Abbas Wits QC SS 2019

 Quantum squared-distance classifier

index1 | 0⟩ H RY H

index2 | 0⟩ H RY

index3 | 0⟩ H
RY
label | 0⟩ X X X X

Conditional
measurement

Amira Abbas Wits QC SS 2019

But why on earth would we do this?
 QML Algorithms
In general…

| 0⟩
| 0⟩ Do something
U(x) ⟨Observable⟩
⋮ else
| 0⟩
Prepare some Encode data Other
Measure
initial state operations

Amira Abbas Wits QC SS 2019

 Quantum-classical machine learning

● Active research to find a design/idea that is useful


● There is no advantage to the toy example shown, but can we

perhaps find a quantum algorithm that is of use?

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Eigenvalue solver - to find eigenvalues/vectors of a local Hamiltonian

Classical Quantum

O(N 2) O((logN )4)

Can actually run on NISQ devices!

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

● Eigenvalue/vector problem


● Classical optimisation (borrowed from classical machine learning)


● Variational circuits 


● Quantum gradients

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

● Eigenvalue/vector problem


● Classical optimisation (borrowed from classical machine learning)


● Variational circuits 


● Quantum gradients

Amira Abbas Wits QC SS 2019

 Classical machine learning

Amira Abbas Wits QC SS 2019

 Classical machine learning

f(x) = y

Amira Abbas Wits QC SS 2019

 Classical machine learning

f(x) = y

y
f

Amira Abbas Wits QC SS 2019

 Classical machine learning

f(x) = y
Approximate

y y
f
g(x; θ) = ŷ g

x x

Amira Abbas Wits QC SS 2019

 Classical machine learning

f(x) = y

y y
f
g(x; θ) = ŷ g

x Update θ x

g→f
Amira Abbas Wits QC SS 2019
 Cost/loss function

cost(θ, y,̂ y)

Amira Abbas Wits QC SS 2019

 Cost/loss function

cost(θ, y,̂ y)
f(x) = y g(x; θ) = ŷ

Amira Abbas Wits QC SS 2019

 Cost/loss function

cost(θ, y,̂ y)
f(x) = y g(x; θ) = ŷ
y y
f g

x x

Amira Abbas Wits QC SS 2019

 Cost/loss function

cost(θ, y,̂ y)
f(x) = y g(x; θ) = ŷ
y y
f g

x x

Amira Abbas Wits QC SS 2019

 Cost/loss function

cost(θ*, y,̂ y)
f(x) = y g(x; θ*) = ŷ
y y
f g

x x

Amira Abbas Wits QC SS 2019

How do we update θ ?
 Optimisation

cost(θ, y,̂ y)

Amira Abbas Wits QC SS 2019

 Optimisation

cost(θ, y,̂ y)

cost

θ
Amira Abbas Wits QC SS 2019
 Optimisation

cost

θ* θ
Amira Abbas Wits QC SS 2019
 Gradient-based optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Gradient-based optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Gradient-based optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Gradient-based optimisation

cost

θ2 θ1 θ
Amira Abbas Wits QC SS 2019
 Gradient-based optimisation

cost

θ2 θ1 θ
Amira Abbas Wits QC SS 2019
 Optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Optimisation

cost

θ1 θ
Amira Abbas Wits QC SS 2019
 Optimisation

● Want to minimise some cost function w.r.t. parameters of your model

Amira Abbas Wits QC SS 2019

 Variational circuits

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩
| 0⟩

Prepare some
initial state

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩
| 0⟩

Prepare some Gate

initial state operations

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩ U ⟨S⟩
| 0⟩

Prepare some Gate

initial state operations

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩ U(θ) ⟨Sθ⟩ = g(x; θ)
| 0⟩

Prepare some Gate

initial state operations

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩ U(θ) ⟨Sθ⟩ = g(x; θ)
| 0⟩

Prepare some Gate

initial state operations

Cost fn

Amira Abbas Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩ U(θ) ⟨Sθ⟩ = g(x; θ)
| 0⟩

Prepare some Gate

initial state operations

Cost fn

∇θ Sθ
Amira Abbas Wits QC SS 2019
 Variational circuits

| 0⟩
| 0⟩ U(θ) ⟨Sθ⟩ = g(x; θ)
| 0⟩

Prepare some Gate

initial state operations

Update θ Cost fn

∇θ Sθ
Amira Abbas Wits QC SS 2019
 Variational circuits

| 0⟩
| 0⟩ U(θ) ⟨Sθ⟩ = g(x; θ)
| 0⟩

Prepare some Gate

initial state operations

Update θ Cost fn

∇θ Sθ
Amira Abbas Wits QC SS 2019
 Problem?

Amira Abbas Wits QC SS 2019

 Quantum gradients?

Amira Abbas Wits QC SS 2019

 Quantum gradients

| 0⟩ | 0⟩
∇θ Sθ = | 0⟩ U(θ + h) − | 0⟩ U(θ − h)
| 0⟩ | 0⟩

Amira Abbas See https://pennylane.ai/qml/concepts/autograd_quantum.html for technical proof Wits QC SS 2019

 Quantum gradients

| 0⟩ | 0⟩
∇θ Sθ = | 0⟩ U(θ + h) − | 0⟩ U(θ − h)
| 0⟩ | 0⟩

∇θ Sθ = ⟨Sθ+h⟩ − ⟨Sθ−h⟩

Amira Abbas See https://pennylane.ai/qml/concepts/autograd_quantum.html for technical proof Wits QC SS 2019

 Variational circuits

| 0⟩
| 0⟩ U(θ*) ⟨Sθ*⟩
| 0⟩

Prepare some Gate

initial state operations

Update θ Cost fn

Amira Abbas Wits QC SS 2019

Putting this together with VQE
 Variational Quantum Eigensolver (VQE)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

3 ● Solving a colouring puzzle


3 2 ● Assume we can only colour nodes in red

or blue

1
● Edge weights are counted if they connect
two different colour nodes


1 3 ● How can we colour this such that the

score is maximised? 


● Hint: 12

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

3 Score: 3 + 1 + 3 = 7

3 2

1 3

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

3 MaxCut solution: 3 + 2 + 3 + 1 + 3 = 12

3 2

1 3

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

C
2
1
A
1 B

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

C
2
1
A
1 B

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

23 = 8 possibilities

C
2
1
A
1 B

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

CBA
3
2 = 8 possibilities | 000⟩
| 001⟩
C | 010⟩
2
| 011⟩
1 | 100⟩
A
| 101⟩
1 B | 110⟩
| 111⟩

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Ĥ CBA
0000000000000000000000000000000
2 0 0 0 0 0 0 0 | 000⟩
0 −1 0 0 0 0 0 0 0 | 001⟩
0 0 0 0 0 0 0 0 0 | 010⟩
0 0 0 −1 0 0 0 0 0 | 011⟩
0 0 0 0 −1 0 0 0 0 | 100⟩
0 0 0 0 0 0 0 0 0 | 101⟩
0 0 0 0 0 0 −1 0 0 | 110⟩
0 0 0 0 0 0 0 2 0 | 111⟩

https://www.scottaaronson.com/qclec/25.pdf

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Ĥ | 000⟩
0000000000000000000000000000000
2 0 0 0 0 0 0 0 1
0 −1 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 −1 0 0 0 0 0 0
0 0 0 0 −1 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 −1 0 0 0
0 0 0 0 0 0 0 2 0 0

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Ĥ | 000⟩
0000000000000000000000000000000
2 0 0 0 0 0 0 0 1 2
0 −1 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 −1 0 0 0 0 0 0 0
0 0 0 0 −1 0 0 0 0 0 = 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 −1 0 0 0 0
0 0 0 0 0 0 0 2 0 0 0

Ĥ | 000⟩

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

Ĥ | 011⟩
0000000000000000000000000000000
2 0 0 0 0 0 0 0 0 0
0 −1 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 −1 0 0 0 0 0 1 −1
0 0 0 0 −1 0 0 0 0 0 = 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 −1 0 0 0 0
0 0 0 0 0 0 0 2 0 0 0

Ĥ | 011⟩

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

⟨011 | Ĥ | 011⟩

0
0000000000000000000000000000000
2 0 0 0 0 0 0 0
0 −1 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
0 0 0 −1 0 0 0 0 0 1
[00010000] 0 0 0 0 −1 0 0 0 0 0 = −1
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 −1 0 0 0
0 0 0 0 0 0 0 2 0 0

⟨011 | Ĥ | 011⟩

Amira Abbas Source: https://medium.com/qiskit/the-variational-quantum-eigensolver-43f7718c2747 Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| ψ⟩ U(θ) ⟨ψ | Ĥ | ψ⟩

Update θ Cost fn

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H

| 0⟩ H

| 0⟩ H

| 000⟩
| 001⟩
| 010⟩
| 011⟩
| ψ⟩ →
| 100⟩
| 101⟩
| 110⟩
Amira Abbas | 111⟩ Wits QC SS 2019
 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2)

| 0⟩ H Ry(θ3)

(Made up example)
| ψ⟩

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2)

| 0⟩ H Ry(θ3)

| ψ⟩ Ĥ | ψ⟩

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2) ⟨ψ | Ĥ | ψ⟩
| 0⟩ H Ry(θ3)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2) ⟨ψ | Ĥ | ψ⟩
| 0⟩ H Ry(θ3)

̂
cost(θi, ⟨H⟩)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2) ⟨ψ | Ĥ | ψ⟩
| 0⟩ H Ry(θ3)

Update θi ̂
cost(θi, ⟨H⟩)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 0⟩ H Ry(θ1)

| 0⟩ H Ry(θ2) ⟨ψ | Ĥ | ψ⟩
| 0⟩ H Ry(θ3)

Update θi ̂
cost(θi, ⟨H⟩)

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 000⟩
| 001⟩
| 010⟩
| 011⟩
| ψ⟩ →
| 100⟩
⟨ψ | Ĥ | ψ⟩ ̂
cost(θi, ⟨H⟩)
| 101⟩
| 110⟩
| 111⟩

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)

| 000⟩
| 001⟩
| 010⟩ C
2
| 011⟩
| ψ⟩ → 1 ̂
cost(θi, ⟨H⟩)
| 100⟩ A
| 101⟩
1 B
| 110⟩
| 111⟩

Amira Abbas Wits QC SS 2019

 Variational Quantum Eigensolver (VQE)
The code

Source:https://github.com/Qiskit/qiskit-iqx-tutorials/blob/master/qiskit/advanced/aqua/optimization/max_cut_and_tsp.ipynb

Amira Abbas Wits QC SS 2019

 Resources
• VQE paper: Peruzzo, Alberto, et al. "A variational eigenvalue solver on a photonic
quantum processor." Nature communications 5 (2014): 4213.


• Kandala, Abhinav, et al. "Hardware-efficient variational quantum eigensolver for small

molecules and quantum magnets." Nature 549.7671 (2017): 242.


• VQE Blog (James Weaver): https://medium.com/qiskit/the-variational-quantum-

eigensolver-43f7718c2747


• QML Book: Schuld, Maria, and Francesco Petruccione. Supervised Learning with
Quantum Computers. Vol. 17. Springer, 2018.


• Library of quantum algorithms: https://quantumalgorithmzoo.org/


• EdX course: Quantum Machine Learning. Peter Wittek 2019


• Qiskit Aqua Github: https://github.com/Qiskit/qiskit-aqua/tree/master/qiskit/aqua/

algorithms/adaptive/vqe
Amira Abbas Wits QC SS 2019