← Library · Core concept
Gradient Descent
Gradient Descent is an optimization algorithm used to minimize the loss function of an AI model by iteratively adjusting its internal parameters (weights and biases). It calculates the 'gradient' or slope of the loss function, indicating the direction of steepest descent, and then takes small steps in that direction until a minimum is reached. This process helps the model find the best parameter values.
In plain terms
Gradient Descent is like walking down a foggy hill by always taking a small step in the steepest downward direction until you reach the bottom.
Why it matters
It is the fundamental mathematical engine that allows neural networks and many other AI models to learn from data.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free