In this demo, we will build an optimized lead scoring model using EvalML. To optimize the pipeline, we will set up an objective function to maximize the revenue generated with true positives while taking into account the cost of false positives. At the end of this demo, we also show you how introducing the right objective during the training is over 6x better than using a generic machine learning metric like AUC.
[1]:
import evalml from evalml import AutoMLSearch from evalml.objectives import LeadScoring
To optimize the pipelines toward the specific business needs of this model, you can set your own assumptions for how much value is gained through true positives and the cost associated with false positives. These parameters are
true_positive - dollar amount to be gained with a successful lead
true_positive
false_positive - dollar amount to be lost with an unsuccessful lead
false_positive
Using these parameters, EvalML builds a pileline that will maximize the amount of revenue per lead generated.
[2]:
lead_scoring_objective = LeadScoring( true_positives=1000, false_positives=-10 )
We will be utilizing a dataset detailing a customer’s job, country, state, zip, online action, the dollar amount of that action and whether they were a successful lead.
[3]:
from urllib.request import urlopen import pandas as pd import woodwork as ww customers_data = urlopen('https://featurelabs-static.s3.amazonaws.com/lead_scoring_ml_apps/customers.csv') interactions_data = urlopen('https://featurelabs-static.s3.amazonaws.com/lead_scoring_ml_apps/interactions.csv') leads_data = urlopen('https://featurelabs-static.s3.amazonaws.com/lead_scoring_ml_apps/previous_leads.csv') customers = pd.read_csv(customers_data) interactions = pd.read_csv(interactions_data) leads = pd.read_csv(leads_data) X = customers.merge(interactions, on='customer_id').merge(leads, on='customer_id') y = X['label'] X = X.drop(['customer_id', 'date_registered', 'birthday','phone', 'email', 'owner', 'company', 'id', 'time_x', 'session', 'referrer', 'time_y', 'label', 'country'], axis=1) display(X.head())
We will convert our data into Woodwork data structures. Doing so enables us to have more control over the types passed to and inferred by AutoML.
[4]:
X = ww.DataTable(X, semantic_tags={'job': 'category'}, logical_types={'job': 'Categorical'}) y = ww.DataColumn(y) X.types
In order to validate the results of the pipeline creation and optimization process, we will save some of our data as a holdout set.
EvalML natively supports one-hot encoding and imputation so the above NaN and categorical values will be taken care of.
NaN
[5]:
X_train, X_holdout, y_train, y_holdout = evalml.preprocessing.split_data(X, y, problem_type='binary', test_size=0.2, random_state=0) print(X.types)
Physical Type Logical Type Semantic Tag(s) Data Column job category Categorical ['category'] state category Categorical ['category'] zip float64 Double ['numeric'] action category Categorical ['category'] amount float64 Double ['numeric']
Because the lead scoring labels are binary, we will use AutoMLSearch(X_train=X_train, y_train=y_train, problem_type='binary'). When we call .search(), the search for the best pipeline will begin.
AutoMLSearch(X_train=X_train, y_train=y_train, problem_type='binary')
.search()
[6]:
automl = AutoMLSearch(X_train=X_train, y_train=y_train, problem_type='binary', objective=lead_scoring_objective, additional_objectives=['auc'], max_batches=1, optimize_thresholds=True) automl.search()
{'message': 'The following labels fall below 10% of the target: [True]', 'data_check_name': 'ClassImbalanceDataCheck', 'level': 'warning', 'code': 'CLASS_IMBALANCE_BELOW_THRESHOLD', 'details': {'target_values': [True]}} Generating pipelines to search over... ***************************** * Beginning pipeline search * ***************************** Optimizing for Lead Scoring. Greater score is better. Searching up to 1 batches for a total of 9 pipelines. Allowed model families: extra_trees, linear_model, xgboost, decision_tree, lightgbm, random_forest, catboost
Batch 1: (1/9) Mode Baseline Binary Classification P... Elapsed:00:00 Starting cross validation Finished cross validation - mean Lead Scoring: 0.000 Batch 1: (2/9) Logistic Regression Classifier w/ Imp... Elapsed:00:01 Starting cross validation Finished cross validation - mean Lead Scoring: 13.784 High coefficient of variation (cv >= 0.2) within cross validation scores. Logistic Regression Classifier w/ Imputer + One Hot Encoder + Standard Scaler may not perform as estimated on unseen data. Batch 1: (3/9) Random Forest Classifier w/ Imputer +... Elapsed:00:04 Starting cross validation Finished cross validation - mean Lead Scoring: 12.062 High coefficient of variation (cv >= 0.2) within cross validation scores. Random Forest Classifier w/ Imputer + One Hot Encoder may not perform as estimated on unseen data. Batch 1: (4/9) XGBoost Classifier w/ Imputer + One H... Elapsed:00:06 Starting cross validation Finished cross validation - mean Lead Scoring: 13.173 Batch 1: (5/9) CatBoost Classifier w/ Imputer Elapsed:00:08 Starting cross validation Finished cross validation - mean Lead Scoring: 12.640 High coefficient of variation (cv >= 0.2) within cross validation scores. CatBoost Classifier w/ Imputer may not perform as estimated on unseen data. Batch 1: (6/9) Elastic Net Classifier w/ Imputer + O... Elapsed:00:09 Starting cross validation Finished cross validation - mean Lead Scoring: 15.997 Batch 1: (7/9) Extra Trees Classifier w/ Imputer + O... Elapsed:00:11 Starting cross validation Finished cross validation - mean Lead Scoring: 13.028 Batch 1: (8/9) LightGBM Classifier w/ Imputer + One ... Elapsed:00:13 Starting cross validation Finished cross validation - mean Lead Scoring: 14.635 Batch 1: (9/9) Decision Tree Classifier w/ Imputer +... Elapsed:00:14 Starting cross validation Finished cross validation - mean Lead Scoring: 11.531 Search finished after 00:16 Best pipeline: Elastic Net Classifier w/ Imputer + One Hot Encoder + Standard Scaler Best pipeline Lead Scoring: 15.997325
Once the fitting process is done, we can see all of the pipelines that were searched, ranked by their score on the lead scoring objective we defined.
[7]:
automl.rankings
To select the best pipeline we can call automl.best_pipeline.
automl.best_pipeline
[8]:
best_pipeline = automl.best_pipeline
You can get more details about any pipeline, including how it performed on other objective functions by calling .describe_pipeline() and specifying the id of the pipeline.
.describe_pipeline()
id
[9]:
automl.describe_pipeline(automl.rankings.iloc[0]["id"])
************************************************************************* * Elastic Net Classifier w/ Imputer + One Hot Encoder + Standard Scaler * ************************************************************************* Problem Type: binary Model Family: Linear Pipeline Steps ============== 1. Imputer * categorical_impute_strategy : most_frequent * numeric_impute_strategy : mean * categorical_fill_value : None * numeric_fill_value : None 2. One Hot Encoder * top_n : 10 * features_to_encode : None * categories : None * drop : None * handle_unknown : ignore * handle_missing : error 3. Standard Scaler 4. Elastic Net Classifier * alpha : 0.5 * l1_ratio : 0.5 * n_jobs : -1 * max_iter : 1000 * penalty : elasticnet * loss : log Training ======== Training for binary problems. Objective to optimize binary classification pipeline thresholds for: <evalml.objectives.lead_scoring.LeadScoring object at 0x7f19c91892b0> Total training time (including CV): 1.7 seconds Cross Validation ---------------- Lead Scoring AUC # Training # Validation 0 12.374 0.500 2479.000 1550.000 1 18.058 0.500 2479.000 1550.000 2 17.560 0.500 2480.000 1549.000 mean 15.997 0.500 - - std 3.148 0.000 - - coef of var 0.197 0.000 - -
Finally, we retrain the best pipeline on all of the training data and evaluate it on the holdout dataset.
[10]:
best_pipeline.fit(X_train, y_train)
GeneratedPipeline(parameters={'Imputer':{'categorical_impute_strategy': 'most_frequent', 'numeric_impute_strategy': 'mean', 'categorical_fill_value': None, 'numeric_fill_value': None}, 'One Hot Encoder':{'top_n': 10, 'features_to_encode': None, 'categories': None, 'drop': None, 'handle_unknown': 'ignore', 'handle_missing': 'error'}, 'Elastic Net Classifier':{'alpha': 0.5, 'l1_ratio': 0.5, 'n_jobs': -1, 'max_iter': 1000, 'penalty': 'elasticnet', 'loss': 'log'},})
Now, we can score the pipeline on the holdout data using both the lead scoring score and the AUC.
[11]:
best_pipeline.score(X_holdout, y_holdout, objectives=["auc", lead_scoring_objective])
OrderedDict([('AUC', 0.5), ('Lead Scoring', 15.382631126397248)])
To demonstrate the importance of optimizing for the right objective, let’s search for another pipeline using AUC, a common machine learning metric. After that, we will score the holdout data using the lead scoring objective to see how the best pipelines compare.
[12]:
automl_auc = evalml.AutoMLSearch(X_train=X_train, y_train=y_train, problem_type='binary', objective='auc', additional_objectives=[], max_batches=1, optimize_thresholds=True) automl_auc.search()
{'message': 'The following labels fall below 10% of the target: [True]', 'data_check_name': 'ClassImbalanceDataCheck', 'level': 'warning', 'code': 'CLASS_IMBALANCE_BELOW_THRESHOLD', 'details': {'target_values': [True]}} Generating pipelines to search over... ***************************** * Beginning pipeline search * ***************************** Optimizing for AUC. Greater score is better. Searching up to 1 batches for a total of 9 pipelines. Allowed model families: extra_trees, linear_model, xgboost, decision_tree, lightgbm, random_forest, catboost
Batch 1: (1/9) Mode Baseline Binary Classification P... Elapsed:00:00 Starting cross validation Finished cross validation - mean AUC: 0.500 Batch 1: (2/9) Logistic Regression Classifier w/ Imp... Elapsed:00:00 Starting cross validation Finished cross validation - mean AUC: 0.695 Batch 1: (3/9) Random Forest Classifier w/ Imputer +... Elapsed:00:00 Starting cross validation Finished cross validation - mean AUC: 0.702 Batch 1: (4/9) XGBoost Classifier w/ Imputer + One H... Elapsed:00:01 Starting cross validation Finished cross validation - mean AUC: 0.709 Batch 1: (5/9) CatBoost Classifier w/ Imputer Elapsed:00:03 Starting cross validation Finished cross validation - mean AUC: 0.557 Batch 1: (6/9) Elastic Net Classifier w/ Imputer + O... Elapsed:00:03 Starting cross validation Finished cross validation - mean AUC: 0.500 Batch 1: (7/9) Extra Trees Classifier w/ Imputer + O... Elapsed:00:04 Starting cross validation Finished cross validation - mean AUC: 0.722 Batch 1: (8/9) LightGBM Classifier w/ Imputer + One ... Elapsed:00:05 Starting cross validation Finished cross validation - mean AUC: 0.678 Batch 1: (9/9) Decision Tree Classifier w/ Imputer +... Elapsed:00:05 Starting cross validation Finished cross validation - mean AUC: 0.587 Search finished after 00:06 Best pipeline: Extra Trees Classifier w/ Imputer + One Hot Encoder Best pipeline AUC: 0.722008
[13]:
automl_auc.rankings
Like before, we can look at the rankings and pick the best pipeline.
[14]:
best_pipeline_auc = automl_auc.best_pipeline # train on the full training data best_pipeline_auc.fit(X_train, y_train)
GeneratedPipeline(parameters={'Imputer':{'categorical_impute_strategy': 'most_frequent', 'numeric_impute_strategy': 'mean', 'categorical_fill_value': None, 'numeric_fill_value': None}, 'One Hot Encoder':{'top_n': 10, 'features_to_encode': None, 'categories': None, 'drop': None, 'handle_unknown': 'ignore', 'handle_missing': 'error'}, 'Extra Trees Classifier':{'n_estimators': 100, 'max_features': 'auto', 'max_depth': 6, 'min_samples_split': 2, 'min_weight_fraction_leaf': 0.0, 'n_jobs': -1},})
[15]:
# get the auc and lead scoring score on holdout data best_pipeline_auc.score(X_holdout, y_holdout, objectives=["auc", lead_scoring_objective])
OrderedDict([('AUC', 0.6800619522514355), ('Lead Scoring', -0.008598452278589854)])
When we optimize for AUC, we can see that the AUC score from this pipeline is better than the AUC score from the pipeline optimized for lead scoring. However, the revenue per lead gained was only $7 per lead when optimized for AUC and was $45 when optimized for lead scoring. As a result, we would gain up to 6x the amount of revenue if we optimized for lead scoring.
$7
$45
This happens because optimizing for AUC does not take into account the user-specified true_positive (dollar amount to be gained with a successful lead) and false_positive (dollar amount to be lost with an unsuccessful lead) values. Thus, the best pipelines may produce the highest AUC but may not actually generate the most revenue through lead scoring.
This example highlights how performance in the real world can diverge greatly from machine learning metrics.