Training a new model

To train a new model to categorise patient feedback text, labelled data is required. Discussions are currently underway to enable the release of the data that the multilabel models in pxtextmining are trained on.

This page breaks down the steps in the function pxtextmining.pipelines.run_sklearn_pipeline, which outputs trained sklearn models. This is a high-level explanation of the processes; for more detailed technical information please see the relevant code reference pages for each function.

# Step 1: Generate a random_state which is used for the train_test_split.
# This means that the pipeline and evaluation should be reproducible.
random_state = random.randint(1,999)

# Step 2: Load the data and isolate the target columns from the dataframe.
df = load_multilabel_data(filename = 'datasets/hidden/multilabeldata_2.csv',
                          target = 'major_categories')

# Step 3: Conduct preprocessing: remove punctuation and numbers, clean whitespace and drop empty lines.
# Split into train and test using the random_state above.
X_train, X_test, Y_train, Y_test = process_and_split_data(
                                        df, target = target,
                                        random_state = random_state)

# Step 4: Instantiate a pipeline and hyperparamter grid for each estimator to be tried.
# Conduct a cross-validated randomized search to identify the hyperparameters
  # producing the best results on the validation set.
# For each estimator, returns the pipeline with the best hyperparameters,
  # together with the time taken to search the pipeline.
models, training_times = search_sklearn_pipelines(X_train, Y_train,
                                        models_to_try = models_to_try,
                                        additional_features = additional_features)

# Step 5: Evaluate each pipeline using the test set, comparing predicted values with real values.
# Performance metrics are recorded together with the time taken to search the pipeline.
model_metrics = []
for i in range(len(models)):
    m = models[i]
    t = training_times[i]
    model_metrics.append(get_multilabel_metrics(X_test, Y_test,
                                        random_state = random_state,
                                        labels = target, model_type = 'sklearn',
                                        model = m, training_time = t))

# Step 6: Save the models and performance metrics to the path specified
write_multilabel_models_and_metrics(models,model_metrics,path=path)