• API Reference - SimpleModelComparison

Simple Model Comparison

This notebooks provides an overview for using and understanding simple model comparison check.

Structure:

• What Is the Purpose of the Check?

• Generate Dataset

• Run the check

What Is the Purpose of the Check?

This check compares your current model to a “simple model”, which is a model designed to produce the best performance achievable using very simple rules, such as “always predict the most common class”. The simple model is used as a baseline model; If your model achieves less or similar score to the simple model, this is an indicator of a possible problem with the model (e.g. it wasn’t trained properly).

Using the parameter strategy, you can select the simple model used in the check:

Strategy	Description
prior (default)	The probability vector always contains the empirical class prior distribution (i.e. the class distribution observed in the training set).
most_frequent	The most frequent prediction is predicted. The probability vector is 1 for the most frequent prediction and 0 for the other predictions.
stratified	The predictions are generated by sampling one-hot vectors from a multinomial distribution parametrized by the empirical class prior probabilities.
uniform	Generates predictions uniformly at random from the list of unique classes observed in y, i.e. each class has equal probability.

Similiar to the Simple Model Comparison check, there is no simple model which is more “correct” to use, each gives a different baseline to compare to, and you may experiment with the different types and see how it performs on your data.

This checks applies only to classification datasets.

Generate Dataset

Note

In this example, we use the pytorch version of the mnist dataset and model. In order to run this example using tensorflow, please change the import statements to:

from deepchecks.vision.datasets.classification import mnist_tensorflow as mnist

from deepchecks.vision.checks import SimpleModelComparison
from deepchecks.vision.datasets.classification import mnist_torch as mnist

train_ds = mnist.load_dataset(train=True, object_type='VisionData')
test_ds = mnist.load_dataset(train=False, object_type='VisionData')

Run the check

We will run the check with the prior model type. The check will use the default classification metrics - precision and recall. This can be overridden by providing an alternative scorer using the scorers` parameter.

check = SimpleModelComparison(strategy='stratified')
result = check.run(train_ds, test_ds)

Processing Train Batches:
|     | 0/1 [Time: 00:00]
Processing Train Batches:
|█████| 1/1 [Time: 00:01]
Processing Train Batches:
|█████| 1/1 [Time: 00:01]

Processing Test Batches:
|     | 0/1 [Time: 00:00]

Processing Test Batches:
|█████| 1/1 [Time: 00:04]

Processing Test Batches:
|█████| 1/1 [Time: 00:04]


Computing Check:
|     | 0/1 [Time: 00:00]


Computing Check:
|█████| 1/1 [Time: 00:00]


Computing Check:
|█████| 1/1 [Time: 00:00]

result.show()

To display the results in an IDE like PyCharm, you can use the following code:

#  result.show_in_window()

The result will be displayed in a new window.

Observe the check’s output

We can see in the results that the check calculates the score for each class in the dataset, and compares the scores between our model and the simple model.

In addition to the graphic output, the check also returns a value which includes all of the information that is needed for defining the conditions for validation.

The value is a dataframe that contains the metrics’ values for each class and dataset:

result.value.sort_values(by=['Class', 'Metric']).head(10)

	Model	Metric	Class	Class Name	Number of samples	Value
8	Simple Model	F1	0	0	980	0.088576
10	Perfect Model	F1	0	0	980	1.000000
20	Given Model	F1	0	0	980	0.985316
0	Simple Model	F1	1	1	1135	0.127734
11	Perfect Model	F1	1	1	1135	1.000000
28	Given Model	F1	1	1	1135	0.956600
9	Simple Model	F1	2	2	1032	0.087613
12	Perfect Model	F1	2	2	1032	1.000000
23	Given Model	F1	2	2	1032	0.978334
2	Simple Model	F1	3	3	1010	0.103704

Define a condition

We can define on our check a condition that will validate our model is better than the simple model by a given margin called gain. For classification we check the gain for each class separately and if there is a class that doesn’t pass the defined gain the condition will fail.

The performance gain is the percent of the improved performance out of the “remaining” unattained performance. Its purpose is to reflect the significance of the said improvement. Take for example for a metric between 0 and 1. A change of only 0.03 that takes us from 0.95 to 0.98 is highly significant (especially in an imbalance scenario), but improving from 0.1 to 0.13 is not a great achievement.

The gain is calculated as:

Let’s add a condition to the check and see what happens when it fails:

check = SimpleModelComparison(strategy='stratified')
check.add_condition_gain_greater_than(min_allowed_gain=0.99)
result = check.run(train_ds, test_ds)
result.show()

Processing Train Batches:
|     | 0/1 [Time: 00:00]
Processing Train Batches:
|█████| 1/1 [Time: 00:01]
Processing Train Batches:
|█████| 1/1 [Time: 00:01]

Processing Test Batches:
|     | 0/1 [Time: 00:00]

Processing Test Batches:
|█████| 1/1 [Time: 00:01]

Processing Test Batches:
|█████| 1/1 [Time: 00:01]


Computing Check:
|     | 0/1 [Time: 00:00]


Computing Check:
|█████| 1/1 [Time: 00:00]


Computing Check:
|█████| 1/1 [Time: 00:00]

We detected that for several classes our gain did not passed the target gain we defined, therefore it failed.

previous

Mean Average Precision Report

next

Single Dataset Performance