Table of Contents
Regardless of your investment strategy, fluctuations are expected in the financial market. Despite this variance, professional investors try to estimate their overall returns. Risks and returns differ based on investment types and other factors, which impact stability and volatility. To attempt to predict returns, there are many computer-based algorithms and models for financial market trading. Yet, with new techniques and approaches, data science could improve quantitative researchers' ability to forecast an investment's return.
Ubiquant Investment (Beijing) Co., Ltd is a leading domestic quantitative hedge fund based in China. Established in 2012, they rely on international talents in math and computer science along with cutting-edge technology to drive quantitative financial market investment. Overall, Ubiquant is committed to creating long-term stable returns for investors.
The objective of this competition is to build models that can forecast an investment's return rate. The evaluation criteria for this competition uses a mean of pearson's correlation coefficient for each time_id.
Lineage of this repo:
- Exploratory data analysis
- Stress testing modelling techniques for forecasting
- Creating ensembles, blending and stacking using best models
The purpose of creating this log is to keep a track of our progress during this competition and keep all the experiments reproducible.
This section should list any major frameworks/libraries used to bootstrap your project. Leave any add-ons/plugins for the acknowledgements section. Here are a few examples.
This is an example of how you may give instructions on setting up your project locally. To get a local copy up and running follow these simple example steps.
This is an example of how to list things you need to use the software and how to install them.
- npm
npm install npm@latest -g
Below is an example of how you can instruct your audience on installing and setting up your app. This template doesn't rely on any external dependencies or services.
- Get a free API Key at https://example.com
- Clone the repo
git clone https://github.com/your_username_/Project-Name.git
- Install NPM packages
npm install
- Enter your API in
config.jsconst API_KEY = 'ENTER YOUR API';
Use this space to show useful examples of how a project can be used. Additional screenshots, code examples and demos work well in this space. You may also link to more resources.
For more examples, please refer to the Documentation
- Add EDA
- Add a list of credible ideas
- Add data pre-processing pipeline
- Add model training pipeline
- Add a backlog of potential ideas for improvement (over baseline)
- Formulate the problem: can we treat it as time series? [explore LSTM baseline approach]
- Define if we need a separate model for each investment_id (placeholder result: discarded)
- Need to find groups b/w investment_id's
- Create folder structure for the repo (model, data, notebooks/train.ipynb, train.py)
- Mean and std comparison on the target variable across investment_id's
- Plot temporal variation of target value across investment_id's
- Fix pearson evaluation on the validation set: mean pearson corr per time id
- Optimize hyperparameters on the final fold
- Define a validation schedule and use it across model training (5-folds at least)
- Add baseline autoencoder-MLP implementation in PTL
Distributed under the MIT License. See LICENSE.txt for more information.
Your Name - @your_twitter - [email protected]
Project Link: https://github.com/parthpankajtiwary/ubiquant