Quantum computing is highly effective in processing the large amounts of data being generated by AI-based systems.
Quantum computing focuses on quantum-based information processing including cryptography, pattern recognition, strategic decision making under constraints, and many similar sensitive and performance-based applications. Table 1 lists its key uses and the algorithms developed for them.
Table 1: Key uses of quantum computing
| Key task | Representative approach/Algorithm |
| Optimization | Quantum Approximate Optimization Algorithm (QAOA), Quantum Annealing |
| Cryptography | Shor’s Algorithm |
| Simulation | Variational Quantum Eigensolver (VQE) |
| Search | Grover’s Algorithm |
| Machine learning | Quantum SVM, Variational Quantum Circuit (VQC), Quantum Neural Network (QNN) |
Quantum computing can be implemented in a range of domains in which supercomputing-like performance is required.
- AI and ML: Pattern recognition, Quantum machine learning (QML), Machine learning, Recommendation systems, Natural language processing (NLP)
- Autonomous systems: Autonomous vehicle decision making, Optimization
- Aviation: Airline scheduling, Optimization
- Chemistry: Computational chemistry, Process simulation
- Climate science: Weather modelling and climate simulations
- Cybersecurity: Code breaking (Shor’s Algorithm), Cryptanalysis, Cryptography, Quantum Key Distribution (QKD), Quantum communication
- Data science: Big Data analytics, Search and optimization
- Disaster management: Earthquake and disaster prediction
- Electronics: Semiconductor design, Circuit simulation
- Energy: Energy optimization, Smart grid management
- Finance: Financial portfolio optimization, Finance fraud detection, Risk analysis and Monte Carlo simulations, Data sampling
- Healthcare: Drug discovery, Process reaction simulation, Genomics and precision medicine, Healthcare diagnostics, Protein folding analysis
- Industry 4.0 and Industry 5.0: Quantum digital twins and simulation
- IoT: Network optimization, Smart city simulation, Resource optimization
- Logistics: Supply chain management and optimization
- Manufacturing: Process optimization
- Robotics: Path planning, Resource optimization
- Smart cities: Urban planning optimization, Traffic flow management
- Space: Space communication, Satellite communication and optimization, Space mission planning
- Telecommunications: Telecommunication network optimization
- Transportation: Vehicle routing problems and resource allocation
Data science and engineering using quantum computing
Quantum computing and information processing now aids real-time data engineering and analytics — multi-dimensional data can be processed with accurate predictions (Table 2).
Table 2: A comparison of the quantum and classical methods of information processing and analytics
| Analytics problem |
Quantum approach |
Classical method |
| Search | Grover’s Algorithm | Linear Search |
| Optimization | QAOA | GA, ACO, PSO |
| Linear algebra | HHL Algorithm | Matrix Solvers |
| Clustering | Quantum K-Means | K-Means |
| Classification | Quantum SVM | SVM |
| PCA | Quantum PCA | PCA |
It leverages the principles of entanglement, superposition, quantum gates and quantum parallelism. While classical computing systems process information using bits, quantum computers use qubits. In quantum information processing multiple states can exist simultaneously. This gives it the potential to navigate and accelerate high performance computational tasks, specifically for research that involves simulation, optimization and multi-dimensional data processing.
Quantum machine learning (QML)
Quantum machine learning (QML) integrates and makes use of hybrid models that combine machine learning techniques with quantum-based algorithms. Hybrid quantum models like Quantum Neural Networks (QNN) and Variational Quantum Classifiers (VQC) use high performance quantum algorithms for data preprocessing and optimization, and implement quantum circuits to execute machine learning and deep learning tasks.
Popular frameworks for quantum machine learning, quantum data engineering, and related tasks are:
- PennyLane, https://pennylane.ai/
- Qiskit, https://www.ibm.com/quantum/qiskit
- Qiskit Machine Learning, https://qiskit-community.github.io/qiskit-machine-learning/
- Tensorflow Quantum, https://www.tensorflow.org/quantum
- Google Cirq, https://quantumai.google/cirq
- OpenFermion, https://quantumai.google/software
- ProjectQ, https://projectq.ch/
- sQUlearn, https://squlearn.github.io/
Pennylane: Open source framework for quantum machine learning (QML) and quantum data engineering (QDE)
URL: https://pennylane.ai/
This open source framework is used for quantum data analytics and quantum information processing with a focus on quantum machine learning, quantum AI, and similar applications.
Pennylane can be used for real world AI based and intelligence applications including computer vision and pattern recognition, financial modelling, hybrid AI applications, natural language processing (NLP), optimization problems, quantum chemistry simulations, quantum data science, quantum machine learning, robotics, autonomous systems, etc.
To install Pennylane, use the following code:
$ pip install pennylane
from pennylane import numpy as np
import pennylane as qml
import pandas as pd
# Sample dataset
mydata = pd.DataFrame({
“Age”: [27, 36, 35],
“PurchasedItem”: [0, 1, 1]
})
print(mydata)
# Integration of Quantum simulator
mydev = qml.device(“default.qubit”, wires=1)
@qml.qnode(mydev)
def quantum_featuremap(x):
# Input features encoding
qml.RX(x/50.0, wires=0)
return qml.expval(qml.PauliZ(0))
# Quantum Transformation on Sample Data
mydata[“Quantum_Feature”] = mydata[“Age”].apply(quantum_featuremap)
print(mydata)
OUTPUT
Age PurchasedItem
0 27 0
1 36 1
2 35 1
Age PurchasedItem Quantum_Feature
0 27 0 0.857709
1 36 1 0.751806
2 35 1 0.764842
The Quantum_Feature you get after executing and simulating the quantum circuit presents the quantum encoding. The quantum circuits work as mathematical transformers for data analytics and data engineering
Google Cirq
URL: https://quantumai.google/cirq
Google Cirq is an open source framework that can be used for the simulation of quantum circuits for quantum data engineering and quantum machine learning.
It has a range of features for experimenting and research in quantum machine learning and quantum data science. Cirq helps developers design, implement, simulate and optimize quantum circuits using Python programming.
Quantum computing-based cloud simulators can be used to solve a range of real world problems. IBM Qiskit, Google Cirq, PennyLane and many other frameworks provide quantum circuits that can be simulated on the cloud, Google Colab notebooks or on local devices to get effective and accurate solutions in areas being researched.
