Machine Learning in Elixir: Missing option in confusion matrix (page 104)

Community PragProg Customers
,
1

Version: B1.0

On page 104 the code:

Scholar.Metrics.confusion_matrix(test_targets, test_preds)

throws an error: required :num_classes option not found, received options: []

I think the code should be:

Scholar.Metrics.confusion_matrix(test_targets, test_preds, num_classes: 3)

Another similar error on page 107, the following code fails:

model = Scholar.Neighbors.KNearestNeighbors.fit(train_inputs, train_targets)

I think it should be:

model = Scholar.Neighbors.KNearestNeighbors.fit(train_inputs, train_targets, num_classes: 3)
2

In addition the accuracy I reach is lower ( 0.5555555820465088) and the Confussion matrix:

nx.Tensor<
s64[3][3]
EXLA.Backend<host:0, 0.2777305747.1958084637.177673>
[
[11, 0, 0],
[6, 5, 3],
[7, 0, 4]
]

The dimension of the model is f32[3][13] instead of f32[13][3].

The prediction function also needs the model as input parameter.

3

Thanks for catching these, I’ve updated the book to fix these issues! As far as accuracy, it’s possible it may just be a one-time thing. Is your accuracy consistently lower than what the book states?