Custom metrics in TensorFlow - Model Metrics & Evaluation

Metrics & Evaluation - Custom metrics

Which metric matters for Custom Metrics and WHY

Custom metrics let you measure exactly what matters for your problem. Sometimes built-in metrics like accuracy or loss don't tell the full story. For example, if you want to track how many predictions are above a certain threshold or measure a special score, a custom metric helps. It gives you control to see if your model is really learning what you care about.

Confusion matrix or equivalent visualization

Confusion Matrix Example (for classification):

          Predicted
          Pos   Neg
Actual Pos  TP    FN
       Neg  FP    TN

Custom metrics can also be numeric values like mean squared error or any function you define.

Example: Custom metric tracking how many predictions > 0.5

Predictions: [0.7, 0.2, 0.9, 0.4]
Count > 0.5 = 2
    

Precision vs Recall tradeoff with examples

Custom metrics can help balance tradeoffs like precision and recall by defining exactly what you want to measure.

For example, if you want to track how many positive predictions are correct (precision), you can write a custom metric for that.

Or if you want to track how many actual positives you found (recall), you can create a custom metric for recall.

This helps you understand if your model is better at avoiding false alarms or catching all positives.

What "good" vs "bad" metric values look like for Custom Metrics

Good custom metric values depend on what you measure. For example, if your custom metric counts correct predictions above a threshold, a higher number is good.

Bad values mean your metric shows your model is not doing what you want. For example, if your custom metric measures how often predictions are above 0.5 but the value is very low, your model might be underconfident.

Always compare your custom metric values to a baseline or random guess to know if your model is improving.

Common pitfalls with Custom Metrics

• Defining metrics that are not differentiable or incompatible with training can cause errors.
• Metrics that are too complex or slow can slow down training.
• Confusing metric output with loss: metrics measure performance but do not guide training directly.
• Not resetting metric state between epochs can give wrong results.
• Using metrics that do not match your problem goals can mislead model evaluation.

Self-check: Your model has 98% accuracy but 12% recall on fraud. Is it good?

No, it is not good for fraud detection. The low recall (12%) means your model misses most fraud cases, which is dangerous. Even though accuracy is high, it is misleading because fraud is rare. You want high recall to catch as many fraud cases as possible.

Key Result

Custom metrics let you measure exactly what matters, helping you evaluate model performance beyond standard metrics.

Practice

1. What is the main purpose of a custom metric in TensorFlow model training?

easy

A. To track a special score based on true labels and predictions during training

B. To change the model's architecture automatically

C. To speed up the training process by using GPU

D. To save the model weights after each epoch

Solution

1. Step 1: Understand what metrics do

   Metrics measure how well the model performs by comparing true labels and predictions.

2. Step 2: Identify the role of custom metrics

   Custom metrics let you define your own way to measure performance during training.

3. Final Answer:

   To track a special score based on true labels and predictions during training -> Option A

4. Quick Check:

   Custom metric = special score tracking [OK]

Hint: Custom metrics measure model performance uniquely during training [OK]

Common Mistakes:

• Confusing metrics with model architecture changes
• Thinking metrics speed up training
• Believing metrics save model weights

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

easy

A. def my_metric(y_true, y_pred): print(y_true, y_pred)

B. def my_metric(y_true, y_pred): return np.mean(np.abs(y_true - y_pred))

C. def my_metric(y_true, y_pred): return y_true + y_pred

D. def my_metric(y_true, y_pred): return tf.reduce_mean(tf.abs(y_true - y_pred))

Solution

1. Step 1: Check TensorFlow compatibility

   Custom metrics must use TensorFlow operations like tf.reduce_mean for compatibility.

2. Step 2: Identify correct function

   def my_metric(y_true, y_pred): return tf.reduce_mean(tf.abs(y_true - y_pred)) uses tf.reduce_mean and tf.abs correctly; def my_metric(y_true, y_pred): return np.mean(np.abs(y_true - y_pred)) uses numpy which is incompatible.

3. Final Answer:

   def my_metric(y_true, y_pred): return tf.reduce_mean(tf.abs(y_true - y_pred)) -> Option D

4. Quick Check:

   Use TensorFlow ops in custom metrics [OK]

Hint: Use TensorFlow ops, not numpy, inside custom metric functions [OK]

Common Mistakes:

• Using numpy functions inside metric causing errors
• Returning tensors without reduction to scalar
• Printing values instead of returning metric

3. Given the following custom metric code, what will be the output metric value for y_true = [1, 2, 3] and y_pred = [1, 2, 4]?

def custom_metric(y_true, y_pred):
    diff = tf.abs(y_true - y_pred)
    return tf.reduce_mean(diff)

medium

A. 1.0

B. 0.6667

C. 0.3333

D. 2.0

Solution

1. Step 1: Calculate absolute differences

   Differences: |1-1|=0, |2-2|=0, |3-4|=1

2. Step 2: Compute mean of differences

   Mean = (0 + 0 + 1) / 3 = 0.3333

3. Final Answer:

   0.3333 -> Option C

4. Quick Check:

   Mean absolute difference = 0.3333 [OK]

Hint: Calculate absolute differences, then average them [OK]

Common Mistakes:

• Summing differences without averaging
• Using squared difference instead of absolute
• Mixing up y_true and y_pred order

4. Identify the error in this custom metric function and choose the correct fix:

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

medium

A. Convert y_true and y_pred to numpy arrays first

B. Use tf.reduce_mean instead of tf.reduce_sum to get average error

C. Replace subtraction with addition

D. Add print statements to debug values

Solution

1. Step 1: Understand metric output

   Metrics should return a scalar representing average error, not total sum.

2. Step 2: Fix by averaging

   Replacing tf.reduce_sum with tf.reduce_mean returns average difference, a better metric.

3. Final Answer:

   Use tf.reduce_mean instead of tf.reduce_sum to get average error -> Option B

4. Quick Check:

   Average error metric needs reduce_mean [OK]

Hint: Use reduce_mean for average, not reduce_sum [OK]

Common Mistakes:

• Using reduce_sum causing large values
• Mixing TensorFlow and numpy operations
• Not returning a scalar metric

5. You want to create a custom metric that counts how many predictions differ from true labels by more than 0.5. Which of the following implementations is correct?

hard

A. def count_large_diff(y_true, y_pred):
diff = tf.abs(y_true - y_pred)
return tf.reduce_sum(tf.cast(diff > 0.5, tf.float32))

B. def count_large_diff(y_true, y_pred):
diff = y_true - y_pred
return tf.reduce_mean(diff > 0.5)

C. def count_large_diff(y_true, y_pred):
diff = tf.abs(y_true - y_pred)
return tf.reduce_sum(diff > 0.5)

D. def count_large_diff(y_true, y_pred):
diff = tf.abs(y_true - y_pred)
return tf.reduce_sum(tf.int32(diff > 0.5))

Solution

1. Step 1: Identify correct comparison and casting

   We must compare differences with 0.5 and cast boolean to float for summation.

2. Step 2: Check TensorFlow functions used

   def count_large_diff(y_true, y_pred): diff = tf.abs(y_true - y_pred) return tf.reduce_sum(tf.cast(diff > 0.5, tf.float32)) uses tf.abs, comparison, tf.cast to float32, and tf.reduce_sum correctly.

3. Final Answer:

   def count_large_diff(y_true, y_pred): diff = tf.abs(y_true - y_pred) return tf.reduce_sum(tf.cast(diff > 0.5, tf.float32)) -> Option A

4. Quick Check:

   Cast booleans to float before summing [OK]

Hint: Cast boolean mask to float before summing [OK]

Common Mistakes:

• Not casting boolean mask before summing
• Using reduce_mean instead of reduce_sum
• Using wrong data type cast like tf.int32