Algorithms IBM Quantum Computing Machine Learning Quantum Machine Learning

Bold New Version of IBM’s Qiskit Quantum Development Kit released includes Quantum Machine Learning and New Nature Module

April 7, 2021

One of the most popular Quantum languages and toolsets, Qiskit, now gets a new upgrade to version 025 which includes some new modules such as Nature. The already popular package Aqua gets a replacement which is aimed, not just at Chemistry but Physics too. There are also enhancements to Quantum Machine Learning (QML) which see Qiskit releasing a Machine Learning Module. QML has been perhaps one of the biggest areas of interest from Quantum Computing in application to a number of fields, so no surprise that the Qiskit toolset now sports QML capability. In total there are now four additional modules: Qiskit Nature (which will replace Aqua), FinanceOptimization and Machine Learning.

Qiskit Nature

Qiskit Nature (replacing Aqua) is designed to allow users to to investigate and simulate problems in natural sciences using quantum algorithms. Traditionally only Aqua was available – aimed at just Chemistry. Qiskit Nature’s modularity and its wide composition of different modules ensures its extensive capabilities to address complex problems across more of science and industry such as physics, chemistry and biology.

Qiskit Machine Learning

Qiskit Machine Learning has been released. This new application module builds on top of Qiskit’s current functionality to create and run (parametrised) quantum circuits, evaluate complex observables, and also crucially for machine learning to be able to automatically evaluate the corresponding gradients with respect to circuit parameters.

Qiskit Machine Learning highlights fundamental computational building blocks: Quantum Kernels and Quantum Neural Networks. The aim is allow users to get to a rapid prototype quickly without requiring advanced quantum computing knowledge. However there is also the need to allow it to support the latest innovations in Quantum Machine Learning research and development. For Machine Learning researchers and engineers familiar with PyTorch, they will be happy to learn of the Qiskit Torch (PyTorch) connector which allows users to integrate quantum neural networks direct into PyTorch (Facebook’s Open Source Machine Learning Framework). Qiskit are likely to offer more connectors in the future. There are also some tutorials that can help you understand the new QML module.

Qiskit Finance

Qiskit Finance is the open-source module that can be used for some of the problems manifest in for stock/securities financial problems such as portfolio optimizations.

Learn more about the developments from Qiskit.