D
ensity
B
ased
S
patial
C
lustering of
A
pplications with
N
oise(DBSCAN)
is one of the clustering algorithms which can find clusters in noisy data. It works even on those datasets where K-Means fail to find meaningful clusters. More information about it can be found
here
.
You can learn more about the DBSCAN algorithm in the below video.
import matplotlib.pyplot as plt
# Create Sample data
from sklearn.datasets import make_moons
X, y= make_moons(n_samples=500, shuffle=True, noise=0.1, random_state=20)
plt.scatter(x= X[:,0], y= X[:,1])
The Silhouette Coefficient is calculated using the mean intra-cluster distance (a) and the mean nearest-cluster distance (b) for each sample. The Silhouette Coefficient for a sample is (b – a) / max(a, b). To clarify, b is the distance between a sample and the nearest cluster that the sample is not a part of. Note that Silhouette Coefficient is only defined if number of labels is 2 <= n_labels <= n_samples – 1.
The best value of the Silhouette Coefficient is 1 and the worst value is -1. Values near 0 indicate overlapping clusters. Negative values generally indicate that a sample has been assigned to the wrong cluster
import pandas as pd
from sklearn.metrics import silhouette_score
from sklearn.cluster import DBSCAN
# Defining the list of hyperparameters to try
eps_list=np.arange(start=0.1, stop=0.9, step=0.01)
min_sample_list=np.arange(start=2, stop=5, step=1)
# Creating empty data frame to store the silhouette scores for each trials
silhouette_scores_data=pd.DataFrame()
for eps_trial in eps_list:
for min_sample_trial in min_sample_list:
# Generating DBSAN clusters
db = DBSCAN(eps=eps_trial, min_samples=min_sample_trial)
if(len(np.unique(db.fit_predict(X)))>1):
sil_score=silhouette_score(X, db.fit_predict(X))
else:
continue
trial_parameters="eps:" + str(eps_trial.round(1)) +" min_sample :" + str(min_sample_trial)
silhouette_scores_data=silhouette_scores_data.append(pd.DataFrame(data=[[sil_score,trial_parameters]], columns=["score", "parameters"]))
# Finding out the best hyperparameters with highest Score
silhouette_scores_data.sort_values(by='score', ascending=False).head(1)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
## Finding best values of eps and min_samples
import
numpy
as
np
import
pandas
as
pd
from
sklearn
.
metrics
import
silhouette_score
from
sklearn
.
cluster
import
DBSCAN
# Defining the list of hyperparameters to try
eps_list
=
np
.
arange
(
start
=
0.1
,
stop
=
0.9
,
step
=
0.01
)
min_sample_list
=
np
.
arange
(
start
=
2
,
stop
=
5
,
step
=
1
)
# Creating empty data frame to store the silhouette scores for each trials
silhouette_scores_data
=
pd
.
DataFrame
(
)
for
eps_trial
in
eps_list
:
for
min_sample_trial
in
min_sample_list
:
# Generating DBSAN clusters
db
=
DBSCAN
(
eps
=
eps_trial
,
min_samples
=
min_sample_trial
)
if
(
len
(
np
.
unique
(
db
.
fit_predict
(
X
)
)
)
&
gt
;
1
)
:
sil_score
=
silhouette_score
(
X
,
db
.
fit_predict
(
X
)
)
else
:
continue
trial_parameters
=
"eps:"
+
str
(
eps_trial
.
round
(
1
)
)
+
" min_sample :"
+
str
(
min_sample_trial
)
silhouette_scores_data
=
silhouette_scores_data
.
append
(
pd
.
DataFrame
(
data
=
[
[
sil_score
,
trial_parameters
]
]
,
columns
=
[
"score"
,
"parameters"
]
)
)
# Finding out the best hyperparameters with highest Score
silhouette_scores_data
.
sort_values
(
by
=
'score'
,
ascending
=
False
)
.
head
(
1
)
# DBSCAN Clustering
from sklearn.cluster import DBSCAN
db = DBSCAN(eps=0.18, min_samples=2)
# Plotting the clusters
plt.scatter(x= X[:,0], y= X[:,1], c=db.fit_predict(X))
Lead Data Scientist
Farukh is an innovator in solving industry problems using Artificial intelligence. His expertise is backed with 10 years of industry experience. Being a senior data scientist he is responsible for designing the AI/ML solution to provide maximum gains for the clients. As a thought leader, his focus is on solving the key business problems of the CPG Industry. He has worked across different domains like Telecom, Insurance, and Logistics. He has worked with global tech leaders including Infosys, IBM, and Persistent systems. His passion to teach inspired him to create this website!
