Did you ever wonder which parameters are most impactful when it comes to neural networkings model tuning? I did. That's why I worked on this project. The goal was to develop intuition about neural network parameter tuning (including hyperparameter).
In this project, I use Keras to build a generic neural network class which can be used to test model performance. With the power of Colab GPU, building and tesing neural networks is a pleasure. So let's do this!
The corresponding Medium article can be found here.
It's easy! You just need to click the "Open in Colab" button at the top of the Jupyter Notebook. Nevertheless, please refer to the below Environment section for detailed software requirements.
- Google Colaboratory
- Python: 3.6.7
- scikit-learn: 0.20.2.
- Keras: 2.2.4
- TensorFlow: 1.12.0
- Pandas: 0.22.0
- Numpy: 1.14.6
- Matplotlib: 2.1.2
- Seaborn: 0.7.1
- XGBoost: 0.7.post4
