In this project, I built a neural network from scratch to carry out a prediction problem on a real dataset! I implemented the neural network from scratch by implementing the gradient descent step and backpropagation through the network.
For best experience with managing dependency I advise you install Anconda or miniconda.
Create a virtual environment with conda
conda create --name deep-learning python=3
Activate environment.
conda activate deep-learning
Install dependencies.
pip install -r requirements.txt
Download or clone this predicting-bike-sharing-patterns repository. Launch the app with jupyter-notebook.
jupyter-notebook Your_first_neural_network.py
Run all code cells in the notebook. Optionally have a look at the file my_answers.py
.