Partnering with academic researchers is an important part of our R&D activities, and our way to contribute to the larger online community.
How we can work together
We support academics in the hypothesis generation stage of their research projects.
Typically the workflow has 3 phases:
1) We process the data sets and provide initial results to our academic partners. This may include computational notebooks, code, graphics, statistical analysis.
2) The academic partner undertakes the research and manuscript preparation on its own terms and timeline; we provide commentary and editing on intermediate results when requested.
3) The academic partner is in charge of the journal submission process, although we can provide support in the results dissemination phase (online and at live events).
What we offer
Computing & processing (it will take us less than one week to process the data once we have agreed on the dataset to be analyzed). The output (mathematical expressions) that we provide can be used to generate networks and other objects and constructs.
Access to an online collaboration space that features DOI minting/journal templates/version control/journal submission (This is optional, we can also maintain regular email correspondence if preferred by the academic partner).
We may contribute with selected data sets. We mainly work with behavioral data (financial time series, blockchain and cryptocurrency data, click streams, cybersecurity, quantum science, and, sensor data), but we are open to collaborating in other fields such as neuroscience, social science, policy and law, journalism, and, the arts.
Rapid idea prototyping. You save a considerable amount of time — instead of spending weeks and months elaborating on possible hypotheses from large data sets, you will get a set of driving variables and mathematical expressions (symbolic regression) to start writing right away. We can iterate as necessary to ensure that quality scientific results are achieved.
Ownership. We aim to produce joint publications. You remain the principal investigator. We can engage as actively as you request us.
Lean on-boarding. No NDAs, complicated proposals, or lengthy legal agreements are required from our side (we are ready to begin collaborating on day 1).
Big data ready. The size of the dataset is not an issue — we routinely process tables of millions of rows and thousands of columns.
Data privacy. We specialize in symbolic languages and near-explainable machine learning/artificial intelligence methods. Therefore, we do not necessarily need to know the meaning of the variables in your dataset, to be able to provide data explorations and analysis of the main mathematical properties (variable names can be masked and you can perform data anonymization if needed). However, if you have limitations in terms of disclosure agreements, there are always good public datasets that can be analyzed.