Declarative Deep Learning


Easy-to-use API & UI for the entire machine learning lifecycle


Star on GitHub


framework







experiment_tracker
↳ How does AIQC compare to other experiment trackers?




(play video with Chrome or enable FireFox embedded content)





Reduce data wrangling with a declarative API



>
<














Write 90% less glue-code with structured pipelines


Tabular
(2D)
Sequence
(3D)
Image
(4D)
Classification
(binary, multi)
Quantification
(regression)
Forecasting
(multivariate)


AIQC provides structured protocols that automate data wrangling processes that vary based on: analysis type (e.g. categorize, quantify, generate), data type (e.g. spreadsheet, sequence, image), and data dimensionality (e.g. timepoints per sample).

The DIY approach of patching together custom code and toolsets for each analysis is not maintainable because it places a skillset burden of both data science and software engineering upon a research team.










How do you quality control (QC) your machine learning lifecycle?


Train Validation Test Inference
Prevent evaluation bias with
3-way+ stratification.
Validate the structure
of new samples.
Prevent data leakage by only using preprocessing information derived from the training split/fold. Prevent data drift by
using original preprocessors.
Prevent overfitting by evaluating each
split/ fold of every model
Detect model rot by reevaluating with supervised datasets.
Ensure reproducibility by using a standardized framework
that records the entire workflow.











Automated visualizations for each split & fold of every model


visualizations






AIQC









sensitivity

Conduct what-if analysis to simulate virtual outcomes







ecosystem







Let's get started!


Use Cases & Tutorials