Custom loss functions in TensorFlow - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a custom loss function in machine learning?

A custom loss function is a user-defined way to measure how well a model's predictions match the true answers. It guides the model to learn by telling it how wrong it is.

beginner

How do you define a custom loss function in TensorFlow?

You write a Python function that takes true values and predicted values as inputs and returns a number showing the error. This function is then passed to the model during training.

intermediate

Why might you want to create a custom loss function instead of using built-in ones?

Sometimes built-in loss functions don't fit your problem well. Custom loss lets you focus on what matters most, like penalizing certain mistakes more or combining multiple errors.

beginner

What are the inputs and outputs of a custom loss function in TensorFlow?

Inputs: true labels (y_true) and predicted labels (y_pred). Output: a single number (tensor) representing the loss value.

beginner

Show a simple example of a custom mean absolute error loss function in TensorFlow.

def custom_mae(y_true, y_pred): return tf.reduce_mean(tf.abs(y_true - y_pred)) This calculates the average absolute difference between true and predicted values.

What does a loss function do in machine learning?

A Visualizes the model's architecture

B Generates new data for training

C Measures how wrong the model's predictions are

D Increases the model's size

A loss function calculates the error between predicted and true values to guide model learning.

Which inputs does a TensorFlow custom loss function receive?

A Optimizer and learning rate

B Model weights and biases

C Training data and validation data

D True labels and predicted labels

Custom loss functions take true labels (y_true) and predicted labels (y_pred) as inputs.

Why create a custom loss function?

A To better match the problem's specific needs

B To speed up training by skipping error calculation

C To avoid using TensorFlow's built-in functions

D To reduce the size of the dataset

Custom loss functions help tailor error measurement to the problem's unique requirements.

What should a custom loss function return?

A A tensor representing the loss value

B The model's predictions

C The training dataset

D The model's architecture

The loss function returns a single number (tensor) showing how wrong the predictions are.

Which of these is a valid custom loss function in TensorFlow?

A def loss(x): return x + 1

B def loss(y_true, y_pred): return tf.reduce_mean(tf.square(y_true - y_pred))

C def loss(y_true): return y_true * 2

D def loss(): return 0

A valid loss function takes y_true and y_pred and returns a loss value, like mean squared error.

Explain how to create and use a custom loss function in TensorFlow.

Think about the function inputs, output, and how it connects to model training.

Describe why custom loss functions can improve model training.

Consider the flexibility custom loss functions provide.

Practice

1. What is the main purpose of a custom loss function in TensorFlow?

easy

A. To change the model's architecture layers

B. To generate new training data from existing data

C. To speed up the training process automatically

D. To define a special way to measure how wrong the model's predictions are

Solution

1. Step 1: Understand the role of loss functions

   Loss functions measure how far the model's predictions are from the true values.

2. Step 2: Identify what custom loss functions do

   Custom loss functions let you define your own way to measure this error, which can help in special cases.

3. Final Answer:

   To define a special way to measure how wrong the model's predictions are -> Option D

4. Quick Check:

   Custom loss = custom error measure [OK]

Hint: Loss functions measure error; custom ones define your own error [OK]

Common Mistakes:

• Confusing loss function with model architecture
• Thinking loss speeds up training automatically
• Mixing loss function with data generation

2. Which of the following is the correct way to define a simple custom loss function in TensorFlow using Python?

easy

A. def custom_loss(y_true, y_pred): return tf.add(y_true, y_pred)

B. def custom_loss(y_true, y_pred): return y_true + y_pred

C. def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.square(y_true - y_pred))

D. def custom_loss(y_true, y_pred): return tf.matmul(y_true, y_pred)

Solution

1. Step 1: Check the function signature

   The function must take true and predicted values as inputs.

2. Step 2: Verify the loss calculation

   Loss should measure error, e.g., mean squared error: mean of squared differences.

3. Final Answer:

   def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.square(y_true - y_pred)) -> Option C

4. Quick Check:

   Loss = mean squared error formula [OK]

Hint: Loss = mean of squared differences for regression [OK]

Common Mistakes:

• Returning sum instead of mean
• Adding true and predicted values instead of difference
• Using matrix multiplication instead of error calculation

3. Given the following custom loss function, what will be the output if y_true = [2.0, 3.0] and y_pred = [1.0, 5.0]?

def custom_loss(y_true, y_pred):
    return tf.reduce_sum(tf.abs(y_true - y_pred))

medium

A. 3.0

B. 4.0

C. 5.0

D. 6.0

Solution

1. Step 1: Calculate absolute differences

   abs(2.0 - 1.0) = 1.0 and abs(3.0 - 5.0) = 2.0

2. Step 2: Sum the absolute differences

   1.0 + 2.0 = 3.0

3. Final Answer:

   3.0 -> Option A

4. Quick Check:

   Sum of absolute errors = 3.0 [OK]

Hint: Sum absolute differences for this loss [OK]

Common Mistakes:

• Confusing sum with mean
• Using squared differences instead of absolute
• Mixing up true and predicted values

4. Identify the error in this custom loss function and select the correct fix:

def custom_loss(y_true, y_pred):
    return tf.reduce_mean(y_true - y_pred)

medium

A. Change function to return y_true + y_pred

B. Use absolute difference or squared difference instead of direct subtraction

C. Remove tf.reduce_mean and return raw difference

D. Use tf.matmul instead of subtraction

Solution

1. Step 1: Understand the problem with direct subtraction

   Direct subtraction can give positive and negative values that cancel out, not measuring error properly.

2. Step 2: Correct the loss calculation

   Use absolute value or square the difference to measure error magnitude correctly.

3. Final Answer:

   Use absolute difference or squared difference instead of direct subtraction -> Option B

4. Quick Check:

   Loss must measure error magnitude, not signed difference [OK]

Hint: Loss must be positive; use abs() or square() [OK]

Common Mistakes:

• Returning signed difference causing cancellation
• Removing mean reduces loss meaning
• Using matrix multiplication wrongly

5. You want to create a custom loss function that penalizes under-predictions more than over-predictions. Which of the following custom loss functions achieves this behavior correctly?

hard

A. def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.where(y_pred < y_true, 2 * (y_true - y_pred), y_pred - y_true))

B. def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.abs(y_true - y_pred))

C. def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.square(y_true - y_pred))

D. def custom_loss(y_true, y_pred): return tf.reduce_sum(y_pred - y_true)

Solution

1. Step 1: Understand the requirement

   We want to penalize under-predictions (y_pred < y_true) more than over-predictions.

2. Step 2: Analyze each option

   def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.where(y_pred < y_true, 2 * (y_true - y_pred), y_pred - y_true)) uses tf.where to multiply the error by 2 when y_pred < y_true, increasing penalty for under-predictions. Options B and C treat errors symmetrically. def custom_loss(y_true, y_pred): return tf.reduce_sum(y_pred - y_true) sums signed differences, which is incorrect.

3. Final Answer:

   def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.where(y_pred < y_true, 2 * (y_true - y_pred), y_pred - y_true)) -> Option A

4. Quick Check:

   Conditional penalty for under-prediction = def custom_loss(y_true, y_pred): return tf.reduce_mean(tf.where(y_pred < y_true, 2 * (y_true - y_pred), y_pred - y_true)) [OK]

Hint: Use tf.where to weight errors differently [OK]

Common Mistakes:

• Using symmetric loss ignoring under/over difference
• Summing signed differences causing wrong loss
• Not using conditional logic for penalty