Post 1 companion app · Click each section to explore · Module 1 / Foundations
Click each box to learn what it means and see a materials science example.
🤖 Artificial Intelligence
click to explore
📊 Machine Learning
click to explore
🧠 Deep Learning
click to explore
👆 Click a box above to learn more
Each level of the hierarchy is a subset of the one above it. Deep Learning ⊂ Machine Learning ⊂ Artificial Intelligence.
AI ⊃ ML ⊃ Deep Learning
A single artificial neuron: the basic building block of all neural networks.
Adjust the inputs and weight, see how the output changes in real time.
INPUT CONTROLS
NEURON OUTPUT
–
What this means:
The neuron computes z = w₁·x₁ + w₂·x₂ + b,
then applies an activation function to get the output.
In a full neural network, millions of such neurons are stacked in layers — each learning its own weights from data.
1. What is the correct relationship between AI, Machine Learning, and Deep Learning?
2. In supervised learning for materials science, what serves as the "label"?
3. A model performs perfectly on training data but poorly on new test data. This is called:
4. Which ML type would you use to automatically group crystal structures by similarity, without any labels?
5. What is the main advantage of an ML model over a DFT calculation for property prediction?