Hybrid algorithm for predicting stock prices

qiskit-advocate / qamp-spring-22

Qiskit advocate mentorship program (QAMP) spring 22 cohort (Mar - Jun 2022)

13 stars 1 forks source link

Hybrid algorithm for predicting stock prices #38

Open MaldoAlberto opened 2 years ago

MaldoAlberto commented 2 years ago

Description

The proposal of this project is to generate a quantum circuit that can represent a Quantum RNN and make the connection with a classical model with the purpose of making a prediction of sotck prices. For this purpose we consider as reference the papers ref1, ref2.

Deliverables

Develop the implementation of an RNN in Qiskit. Prepare a draft for possible submission to a journal.

Mentors details

Mentor
- Name: Alberto Maldonado
- GitHub ID: @MaldoAlberto
- What they do: PhD Student

Number of mentees

Type of mentees

Mentee
- Required:
- Interest in Variational quantum Circuit
- Proficiency in Python and Qiskit
- Nice to have:
- Experience using the QML module of Qiskit

robertloredo commented 2 years ago

@MaldoAlberto @HuangJunye perhaps we can combine efforts here as there might be some similarity/overlaps based on project. Please see issue #29 which I created.

robertloredo commented 2 years ago

@anamariarojas123 Hi Ana, here is the other issue I mentioned that might be a good collaborative link. See my message above. :)

SiddharthaMorales commented 2 years ago

Hi, I'm a mentee in this project

HuangJunye commented 2 years ago

@robertloredo @MaldoAlberto @SiddharthaMorales @jamesphysics Is this project now merged with #29 ? If so can you please close one of the issue and upload the presentation slides for checkpoint 1? Thank you.

SiddharthaMorales commented 2 years ago

Sorry about the slides, here they are. Hybrid algorithm for predicting stock prices #38.pdf

@HuangJunye We haven't been able to set up a meeting and talk with james about joining project, could you contact him?

HuangJunye commented 2 years ago

@SiddharthaMorales Can you please provide your checkpoint 2 updates? Instructions are given in the slack channel.

SiddharthaMorales commented 2 years ago

Project Update May 8

We managed to reproduce the famous example of a LSTM neural network: https://towardsdatascience.com/a-quantum-enhanced-lstm-layer-38a8c135dbfa

in qiskit.

Now we will move forward to use different QNN models and not only the expectation value of Z operators.

We will explore different encodings, optimizators and ansatzes.

SiddharthaMorales commented 2 years ago

We presented all results on the final showcase for this qamp-spring. The presentation is attached here.

Hybrid algorithm for predicting stock prices #38 Final.pdf