This collection contains two things: an open source set of Python tools, and a set of computational essays for exploring Artificial Intelligence, Machine Learning, and Robotics. This is a collaborative effort started by the authors, building on almost a century of collective experience in education and research.
The code and essays are designed to require as few computing resources as necessary, while still allowing readers to experience first-hand the topics covered.
- Douglas Blank - Emeritus Professor of Computer Science, Bryn Mawr College; Head of Research at Comet.ml
- Jim Marshall - Professor in the Computer Science Department at Sarah Lawrence College
- Lisa Meeden - Professor in the Computer Science Department at Swarthmore College
Each computational essay is described at Computational Essays.
aitk
is Python package containing the following modules.
-
aitk - top level package
- aitk.robots - for exploring simulated mobile robots, with cameras and sensors
- aitk.algorithms - for exploring algorithms
- aitk.networks - for constructing and visualizing Keras deep learning models
- aitk.utils - for common utilities
If you already have an environment for running Python, and optionally Jupyter Notebooks, you can simply execute this at the command line:
pip install aitk
If you haven't install Jupyter (and are not running in Google's colab), jump down to "Jupyter Installation".
If you are inside a notebook (say on Google's colab):
%pip install aitk --quiet
If you are setting up your own Jupyter Notebook environment on your
own computer, we recommend using miniconda
.
To use miniconda
:
- First install miniconda
- Next, activate your base environment:
source ~/miniconda/bin/activate
- Create a Python 3.8 conda environment:
conda create --name py38 python=3.8
- Activate it:
conda activate py38
You only need to do step 1 once. To get out of conda, back to your regular system:
-
conda deactivate
(will get out of py38) -
conda deactivate
(will get out of base environment)
After activating your conda environment:
-
pip install "aitk[jupyter]"
(installs all of the requirements to run in Jupyter Lab 3.0) -
pip install pandas tensorflow numpy matplotlib tqdm ipycanvas
(some things you might want)
If you want to work in notebooks and jupyter lab:
pip install jupyterlab
jupyter labextension install @jupyter-widgets/jupyterlab-manager ipycanvas
-
jupyter lab
starts it up, opens browser window
For questions and comments, please use https://github.com/ArtificialIntelligenceToolkit/aitk/discussions/