Learn and Visualize Bayesian Network

Used imports:

import bamt.networks as networks
import bamt.preprocessors as pp

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

from sklearn import preprocessing
from sklearn.ensemble import RandomForestClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.tree import DecisionTreeClassifier

from pgmpy.estimators import K2Score

Importing example data:

data = pd.read_csv(r'../Data/real data/vk_data.csv')
data

Choosing a chunk of data:

cols = ['age', 'sex', 'has_pets', 'is_parent', 'relation', 'is_driver', 'tr_per_month', 'median_tr', 'mean_tr']
data = data[cols]
data[['sex', 'has_pets',  'is_parent', 'relation', 'is_driver']] = data[['sex',     'has_pets',     'is_parent', 'relation', 'is_driver']].astype(str)

Preprocessing data, encode categorical features and discretize numerical features, initialize BN and add nodes:

encoder = preprocessing.LabelEncoder()
discretizer = preprocessing.KBinsDiscretizer(n_bins=5, encode='ordinal', strategy='quantile')

p = pp.Preprocessor([('encoder', encoder), ('discretizer', discretizer)])
discretized_data, est = p.apply(data)

bn = networks.HybridBN(has_logit=True, use_mixture=True) # init BN
info = p.info
info
bn.add_nodes(info)

Learning BN structure and parameters with HillClimbing algorithm:

bn.add_edges(discretized_data,  scoring_function=('K2',K2Score))
bn.set_classifiers(classifiers={'age': DecisionTreeClassifier(),
                             'relation': RandomForestClassifier(),
                             'is_driver': KNeighborsClassifier(n_neighbors=2)})
bn.fit_parameters(data)

Visualize BN structure:

bn.plot('bn.html')

The visualized BN structure will not be rendered by jupyter notebook, but you can see it in the root directory of the project.