Underfitting happens when a machine learning model is too simple to capture patterns in the data.
It performs poorly on both training and test data because it fails to learn the underlying relationships.
Think of it as a student not studying enough to understand the topic.
Causes of Underfitting
- Too simple model (few parameters)
- Insufficient features to represent data
- Over-regularization
- Not enough training
Signs of Underfitting
- Low accuracy on training and test sets
- Model predictions are consistently wrong
- High bias in model behavior
How to Prevent Underfitting
- Use a more complex model (more layers, neurons, or parameters)
- Add more relevant features
- Reduce regularization
- Train longer with sufficient data
Real-World Examples
- Predicting house prices using only one feature (like size)
- Classifying images with a tiny neural network
- Stock predictions with oversimplified models
Conclusion
Underfitting prevents the model from learning data patterns properly. Balancing model complexity is key to achieving accurate machine learning results.
Citations
https://savanka.com/category/learn/ai-and-ml/
https://www.w3schools.com/ai/