Building a Lead Scoring Model with EvalML#

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 significantly better than using a generic machine learning metric like AUC.

[1]:

import evalml
from evalml import AutoMLSearch
from evalml.objectives import LeadScoring

Using `tqdm.autonotebook.tqdm` in notebook mode. Use `tqdm.tqdm` instead to force console mode (e.g. in jupyter console)

Configure 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
false_positive - dollar amount to be lost with an unsuccessful lead

Using these parameters, EvalML builds a pileline that will maximize the amount of revenue per lead generated.

[2]:

lead_scoring_objective = LeadScoring(true_positives=100, false_positives=-5)

Dataset#

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())

	job	state	zip	action	amount
0	Engineer, mining	NY	60091.0	page_view	NaN
1	Psychologist, forensic	CA	NaN	purchase	135.23
2	Psychologist, forensic	CA	NaN	page_view	NaN
3	Air cabin crew	NaN	60091.0	download	NaN
4	Air cabin crew	NaN	60091.0	page_view	NaN

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.init(semantic_tags={"job": "category"}, logical_types={"job": "Categorical"})
y = ww.init_series(y)
X.ww

[4]:

	Physical Type	Logical Type	Semantic Tag(s)
Column
job	category	Categorical	['category']
state	category	Categorical	['category']
zip	Int64	IntegerNullable	['numeric']
action	category	Categorical	['category']
amount	float64	Double	['numeric']

Search for the best pipeline#

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.

[5]:

X_train, X_holdout, y_train, y_holdout = evalml.preprocessing.split_data(
    X, y, problem_type="binary", test_size=0.2, random_seed=0
)

X.ww

[5]:

	Physical Type	Logical Type	Semantic Tag(s)
Column
job	category	Categorical	['category']
state	category	Categorical	['category']
zip	Int64	IntegerNullable	['numeric']
action	category	Categorical	['category']
amount	float64	Double	['numeric']

Because the lead scoring labels are binary, we will use set the problem type to “binary”. When we call .search(), the search for the best pipeline will begin.

[6]:

automl = AutoMLSearch(
    X_train=X_train,
    y_train=y_train,
    problem_type="binary",
    objective=lead_scoring_objective,
    additional_objectives=["auc"],
    allowed_model_families=["extra_trees", "linear_model"],
    max_batches=3,
    verbose=True,
)

automl.search(interactive_plot=False)

AutoMLSearch will use mean CV score to rank pipelines.

*****************************
* Beginning pipeline search *
*****************************

Optimizing for Lead Scoring.
Greater score is better.

Using SequentialEngine to train and score pipelines.
Searching up to 3 batches for a total of None pipelines.
Allowed model families:

Evaluating Baseline Pipeline: Mode Baseline Binary Classification Pipeline
Mode Baseline Binary Classification Pipeline:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 0.000

*****************************
* Evaluating Batch Number 1 *
*****************************

Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 1.424

*****************************
* Evaluating Batch Number 2 *
*****************************

Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler + RF Classifier Select From Model:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 1.360

*****************************
* Evaluating Batch Number 3 *
*****************************

Extra Trees Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Oversampler:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 1.213
Elastic Net Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 1.235
Logistic Regression Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler:
        Starting cross validation
        Finished cross validation - mean Lead Scoring: 1.214

Search finished after 21.16 seconds
Best pipeline: Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler
Best pipeline Lead Scoring: 1.423894

[6]:

{1: {'Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler': 3.5771636962890625,
  'Total time of batch': 3.7019460201263428},
 2: {'Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler + RF Classifier Select From Model': 4.277435064315796,
  'Total time of batch': 4.406996726989746},
 3: {'Extra Trees Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Oversampler': 3.5316975116729736,
  'Elastic Net Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler': 3.454920530319214,
  'Logistic Regression Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler': 4.721990346908569,
  'Total time of batch': 12.171614408493042}}

View rankings and select pipeline#

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

[7]:

	id	pipeline_name	search_order	ranking_score	mean_cv_score	standard_deviation_cv_score	percent_better_than_baseline	high_variance_cv	parameters
0	1	Random Forest Classifier w/ Label Encoder + Im...	1	1.423894	1.423894	0.540845	inf	False	{'Label Encoder': {'positive_label': None}, 'I...
1	2	Random Forest Classifier w/ Label Encoder + Im...	2	1.360457	1.360457	0.590666	inf	False	{'Label Encoder': {'positive_label': None}, 'I...
2	4	Elastic Net Classifier w/ Label Encoder + Sele...	4	1.234589	1.234589	0.430687	inf	False	{'Label Encoder': {'positive_label': None}, 'N...
3	5	Logistic Regression Classifier w/ Label Encode...	5	1.214160	1.214160	0.395051	inf	False	{'Label Encoder': {'positive_label': None}, 'N...
4	3	Extra Trees Classifier w/ Label Encoder + Sele...	3	1.213167	1.213167	0.709773	inf	False	{'Label Encoder': {'positive_label': None}, 'N...
5	0	Mode Baseline Binary Classification Pipeline	0	0.000000	0.000000	0.000000	0.0	False	{'Label Encoder': {'positive_label': None}, 'B...

To select the best pipeline we can call automl.best_pipeline.

[8]:

best_pipeline = automl.best_pipeline

Describe 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.

[9]:

automl.describe_pipeline(automl.rankings.iloc[0]["id"])


***************************************************************************************
* Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler *
***************************************************************************************

Problem Type: binary
Model Family: Random Forest

Pipeline Steps
==============
1. Label Encoder
         * positive_label : None
2. Imputer
         * categorical_impute_strategy : most_frequent
         * numeric_impute_strategy : mean
         * boolean_impute_strategy : most_frequent
         * categorical_fill_value : None
         * numeric_fill_value : None
         * boolean_fill_value : None
3. One Hot Encoder
         * top_n : 10
         * features_to_encode : None
         * categories : None
         * drop : if_binary
         * handle_unknown : ignore
         * handle_missing : error
4. Oversampler
         * sampling_ratio : 0.25
         * k_neighbors_default : 5
         * n_jobs : -1
         * sampling_ratio_dict : None
         * categorical_features : [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20]
         * k_neighbors : 5
5. Random Forest Classifier
         * n_estimators : 100
         * max_depth : 6
         * n_jobs : -1

Training
========
Training for binary problems.
Objective to optimize binary classification pipeline thresholds for: <evalml.objectives.lead_scoring.LeadScoring object at 0x7ff281f52640>
Total training time (including CV): 3.6 seconds

Cross Validation
----------------
             Lead Scoring   AUC # Training # Validation
0                   2.048 0.720      3,099        1,550
1                   1.116 0.599      3,099        1,550
2                   1.107 0.639      3,100        1,549
mean                1.424 0.652          -            -
std                 0.541 0.062          -            -
coef of var         0.380 0.095          -            -

Evaluate on hold out#

Finally, since the best pipeline was trained on all of the training data, we evaluate it on the holdout dataset.

[10]:

best_pipeline_score = best_pipeline.score(
    X_holdout, y_holdout, objectives=["auc", lead_scoring_objective]
)
best_pipeline_score

[10]:

OrderedDict([('AUC', 0.6356527651858568),
             ('Lead Scoring', 1.3155631986242475)])

Why optimize for a problem-specific objective?#

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.

[11]:

automl_auc = evalml.AutoMLSearch(
    X_train=X_train,
    y_train=y_train,
    problem_type="binary",
    objective="auc",
    additional_objectives=[lead_scoring_objective],
    allowed_model_families=["extra_trees", "linear_model"],
    max_batches=3,
    verbose=True,
)

automl_auc.search(interactive_plot=False)

AutoMLSearch will use mean CV score to rank pipelines.

*****************************
* Beginning pipeline search *
*****************************

Optimizing for AUC.
Greater score is better.

Using SequentialEngine to train and score pipelines.
Searching up to 3 batches for a total of None pipelines.
Allowed model families:

Evaluating Baseline Pipeline: Mode Baseline Binary Classification Pipeline
Mode Baseline Binary Classification Pipeline:
        Starting cross validation
        Finished cross validation - mean AUC: 0.500

*****************************
* Evaluating Batch Number 1 *
*****************************

Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler:
        Starting cross validation
        Finished cross validation - mean AUC: 0.652

*****************************
* Evaluating Batch Number 2 *
*****************************

Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler + RF Classifier Select From Model:
        Starting cross validation
        Finished cross validation - mean AUC: 0.646

*****************************
* Evaluating Batch Number 3 *
*****************************

Extra Trees Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Oversampler:
        Starting cross validation
        Finished cross validation - mean AUC: 0.653
Elastic Net Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler:
        Starting cross validation
        Finished cross validation - mean AUC: 0.645
Logistic Regression Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler:
        Starting cross validation
        Finished cross validation - mean AUC: 0.647

Search finished after 23.66 seconds
Best pipeline: Extra Trees Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Oversampler
Best pipeline AUC: 0.653133

[11]:

{1: {'Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler': 4.445448637008667,
  'Total time of batch': 4.570077180862427},
 2: {'Random Forest Classifier w/ Label Encoder + Imputer + One Hot Encoder + Oversampler + RF Classifier Select From Model': 5.078545570373535,
  'Total time of batch': 5.2058820724487305},
 3: {'Extra Trees Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Oversampler': 4.583144903182983,
  'Elastic Net Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler': 4.195231199264526,
  'Logistic Regression Classifier w/ Label Encoder + Select Columns By Type Transformer + Label Encoder + Imputer + Standard Scaler + Select Columns Transformer + Select Columns Transformer + Label Encoder + Imputer + One Hot Encoder + Standard Scaler + Oversampler': 3.992291212081909,
  'Total time of batch': 13.40848994255066}}

[12]:

automl_auc.rankings

[12]:

	id	pipeline_name	search_order	ranking_score	mean_cv_score	standard_deviation_cv_score	percent_better_than_baseline	high_variance_cv	parameters
0	3	Extra Trees Classifier w/ Label Encoder + Sele...	3	0.653133	0.653133	0.058096	15.313288	False	{'Label Encoder': {'positive_label': None}, 'N...
1	1	Random Forest Classifier w/ Label Encoder + Im...	1	0.652452	0.652452	0.061740	15.245206	False	{'Label Encoder': {'positive_label': None}, 'I...
2	5	Logistic Regression Classifier w/ Label Encode...	5	0.646823	0.646823	0.043723	14.682289	False	{'Label Encoder': {'positive_label': None}, 'N...
3	2	Random Forest Classifier w/ Label Encoder + Im...	2	0.645598	0.645598	0.053493	14.559799	False	{'Label Encoder': {'positive_label': None}, 'I...
4	4	Elastic Net Classifier w/ Label Encoder + Sele...	4	0.645471	0.645471	0.042740	14.547088	False	{'Label Encoder': {'positive_label': None}, 'N...
5	0	Mode Baseline Binary Classification Pipeline	0	0.500000	0.500000	0.000000	0.000000	False	{'Label Encoder': {'positive_label': None}, 'B...

Like before, we can look at the rankings and pick the best pipeline.

[13]:

best_pipeline_auc = automl_auc.best_pipeline

[14]:

# get the AUC and lead scoring score on holdout data
best_pipeline_auc_score = best_pipeline_auc.score(
    X_holdout, y_holdout, objectives=["auc", lead_scoring_objective]
)
best_pipeline_auc_score

[14]:

OrderedDict([('AUC', 0.6407071622846781),
             ('Lead Scoring', 0.21066208082545143)])

[15]:

assert best_pipeline_score["Lead Scoring"] >= best_pipeline_auc_score["Lead Scoring"]
assert best_pipeline_auc_score["Lead Scoring"] >= 0

When we optimize for AUC, we can see that the AUC score from this pipeline is similar to the AUC score from the pipeline optimized for lead scoring. However, the revenue per lead is much smaller per lead when optimized for AUC and was much larger when optimized for lead scoring. As a result, we would have a huge gain on the amount of revenue if we optimized for lead scoring.

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.