AIQC

Results 38 issues of AIQC

## Background `Splitset.make` is where the sample indices that make up the splits are defined. ## Problem When labels are not balanced, the network gets biases toward the majority classes...

> Difficulty: ★☆☆☆☆ ## Background When creating and updating entries in a database, it is good practice the have the fields `created_at` and `updated_at` in each table. ## Problem Right...

feature
good first issue

> Difficulty: ★★☆☆☆ > Blocked by: https://github.com/aiqc/aiqc/issues/5 ## Background A Queue requires a Splitset. https://aiqc.readthedocs.io/en/latest/notebooks/api_low_level.html#9.-Create-a-Batch-of-training-Jobs. ## Problem However, what if AIQC doesn't support the user's data type? Or they just...

feature

> Difficulty: ★☆☆☆☆ ## Background When using a validation split/fold to prevent [evaluation bias](https://aiqc.medium.com/) top tune parameters/ topologies, you're not supposed to look at the holdout set. ## Problem Currently,...

good first issue
refactor

> Difficulty: ★★☆☆☆ ## Background We have the `TrainingCallback` class which has the `Keras.MetricCutoff` subclass for early stopping keras models. This is used in`tests.py` module in the `keras_image_binary` class. there...

good first issue

## Background Before running all of the jobs, you may be wondering if your model is set up correctly in the first place. ## Problem You don't want to attempt...

## Background Splitsets and foldsets divide the sample population into evenly distributed sets. However, in unbalanced datasets or highly folded datasets you can't always guarantee that infrequent classes will be...

## Background There are lots of tool emerging that make instant EDA possible. `Dataset.to_pandas()` lets you fetch the data. `datum.py` has lots of methods for making predefined datasets. ## Problem...

good first issue

> Difficulty: ★☆☆☆☆ ## Background `tests.py` has tests for each: - data type (tabular, image) - analysis type (binary classify, multi label classify, regression) - library (pytorch, keras) The `make_test_batch_*`...

good first issue

> Difficulty: ★☆☆☆☆ > Related to: https://github.com/aiqc/aiqc/issues/9 ## Background AIQC stores data in a SQLite file. It has methods like `def get_db()` `def create_db()` and `def destroy_db()` for working with...

feature
good first issue