Week 9: Hands-on Deep Learning
Lab
Overview
A hands-on lab session on deep learning. You build, train, and evaluate neural networks on a real engineering dataset: concrete compressive strength. The session uses two notebooks:
- Lab 1: a neural network with one node. The smallest possible network: load the data, split and scale it, build a baseline, train a single node, and use it for inference on a new concrete mix.
- Lab 2: a neural network with multiple layers. The same task with hidden layers and an activation function: count the parameters, train the network, and compare it fairly against the baseline and the one-node model.
How to run the labs in Google Colab
- Download the notebook and the dataset from the table below.
- Open colab.research.google.com, choose File, Upload notebook, and select the downloaded notebook.
- Choose Runtime, Run all. When the upload dialog appears, select the dataset file
Concrete Compressive Strength.csvfrom your computer. - Read the text between the cells: it explains what each step does and why. Finish with the exercises at the end of each notebook.
No GPU is needed; everything runs on the CPU in under a minute.
Materials
| Material | Where |
|---|---|
| Lab 1 notebook | lab01_one_node.ipynb |
| Lab 2 notebook | lab02_multilayer_network.ipynb |
| Dataset | Concrete Compressive Strength.csv |
Dataset citation: Yeh, I. (1998). Concrete Compressive Strength [Dataset]. UCI Machine Learning Repository. https://doi.org/10.24432/C5PK67.