Fundamentals GUIDE
Loss Functions
A loss defines the objective used to assess predictions or another training target.
Listen to this guideAI-narrated · ~3 min
On this page3 min read
Overview
Individual loss values are often combined into a scalar for optimization, and the chosen objective is a proxy that may differ from the application's evaluation metric.
Deep Dive
A prediction loss compares outputs with training targets according to a specified rule. Those targets can be labels or values constructed for the learning task; they should not automatically be called unquestionable truth. Implementations may produce per-example losses and then sum or average them for an optimization step. Choosing the rule matters. Mean squared error averages squared residuals, giving large residuals more influence than mean absolute error, which averages their absolute values. For residuals one and three, mean squared error is five, while mean absolute error is two. These are illustrative calculations rather than performance results from a dataset. Classification commonly uses cross-entropy. With a one-hot target and a predicted probability p for its class, the loss is minus log p. A lower probability for the target class gives a larger loss, so confident wrong predictions can incur high loss. The logarithm base determines the units: base-two loss is in bits, while a natural logarithm gives nats. For a target probability of one-quarter, minus log-base-two p is two bits. Gradient-based optimization needs usable derivative information, but differentiability requirements should not be generalized to every learning algorithm. Some losses have nondifferentiable points handled by an appropriate subgradient or surrogate. Accuracy is a discrete evaluation metric and is generally not directly useful as a gradient-training objective. A suitable training loss supplies an optimization signal; final task quality needs evaluation using the relevant metrics.
Real-World Implementation
Using cross-entropy loss to train an email spam classifier that penalizes confident misclassifications
Choosing mean absolute error for house-price prediction so a few extreme mansions do not dominate training
Applying a contrastive loss so a face-recognition model pulls together images of the same person
Engineering a reward-model loss to steer a chatbot toward more helpful and honest responses
The Future of Loss Functions
Objective design can combine prediction terms, regularization, and other task requirements. Comparisons should state the target definition, formula, reduction, units, and evaluation metric. An objective can decrease while an application metric remains unchanged or worsens, so improvement should be measured rather than inferred from training loss alone. Custom losses need examination of the behavior they encourage as well as their numerical and optimization properties.
Putting it into practice
Why it matters
- Clearer decisions. It helps you separate clear technical claims from marketing language.
- Cost and budget. You can ask better implementation questions before spending money or time.
- Team and workflow. Teams with shared understanding make better product, policy, and learning decisions.
Watch out for
- Different teams may use the same term differently, so define scope early.
- Benchmarks can look strong while real-world performance is uneven.
- Ignoring data quality and evaluation plans often creates fragile outcomes.
Next steps
- Start with a plain-language definition of the outcome you need.
- Pick one success metric and one failure condition before testing.
- Run a small pilot with representative data, not a polished demo set.
- Document where Loss Functions helps and where simpler methods are better.
Keep Exploring
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Loss Functions quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Frequently asked questions
What is Loss Functions?
A loss defines the objective used to assess predictions or another training target. Individual loss values are often combined into a scalar for optimization, and the chosen objective is a proxy that may differ from the application's evaluation metric.
For a prediction task, what does its selected loss quantify against the training target?
A loss defines a training objective; it does not automatically measure every aspect of application quality.
Which loss function is most commonly used for classification tasks?
Cross-entropy compares predicted probability distributions to true labels and is the standard choice for classification.
Compared to mean squared error, why might you choose mean absolute error?
Mean absolute error treats all errors proportionally rather than squaring them, so a few extreme outliers have less influence.
Why does gradient-based training need a usable loss derivative or suitable substitute?
This concerns gradient-based training, not a claim that every possible learning algorithm requires a differentiable loss.
How does one-hot class cross-entropy change as probability assigned to the target class becomes very small?
The term minus log p grows as the target-class probability shrinks. This does not guarantee a particular gradient magnitude for every model weight.
Keep learning
Related guides
More guides picked for this topic