I am ready for a long road flight for work with a week- or months-long projects.
Contact us
Cookies
Cookie Settings
Cookies necessary for the correct operation of the site are always enabled.
Other cookies are configurable.
Other cookies are configurable.
Predictive models
Platform
Build interpretable and portable predictive models
Our circadian feature extraction algorithm dramatically facilitates building predictive models and hypothesis testing
Among difficulties on the way to incorporate wearables data and circadian rhythm analysis into your study, uncertainty on what is the optimal device and data format, how to handle missing data, and what is the right way for feature extraction — are the major ones. Our platform offers a ready-to-implement solution. On top of that, our feature extraction algorithm was designed specifically to meet the needs of easily interpretable features for models that can be transferred between your longitudinal data from consumer-grade wearable devices to cross-sectional, but rich in medical, health, and lifestyle labels big data from biobanks. This way your results can be easily corroborated by insights on a larger population-scale dataset:
Harness capabilities of deep learning in your work.
Tune or build brand new Neural Network model to extract circadian rhythm features from wearables time series data export
Among difficulties on the way to incorporate wearables data and circadian rhythm analysis into your study, uncertainty on what is the optimal device and data format, how to handle missing data, and what is the right way for feature extraction — are the major ones. Our platform offers a ready-to-implement solution. On top of that, our feature extraction algorithm was designed specifically to meet the needs of easily interpretable features for models that can be transferred between your longitudinal data from consumer-grade wearable devices to cross-sectional, but rich in medical, health, and lifestyle labels big data from biobanks. This way your results can be easily corroborated by insights on a larger population-scale dataset:
Harness capabilities of deep learning in your work.
Tune or build brand new Neural Network model to extract circadian rhythm features from wearables time series data export
New data sources
Easily integrate new data sources to improve on statistical significance of your findings
Кeplace meta studies by direct application of models
Centralized, transferable data format
Combine two worlds
Apply models to both cross-sectional biobank dataset and longitudinal consumer-wearables data streams
Full picture
Analysis of both historical and prospective longitudinal data
Responsible AI
Models based on meaningful and interpretable features — no black-box approaches
Easily test hypotheses on biobank population-scale big data
Big data
Establish baselines for cohorts based on demographic, lifestyle, socio-economic factors
Built-in baselines
Longitudinal and cross-sectional data