jonathanagustin commited on
Commit
9d4f3bc
1 Parent(s): 6d9d7c6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +222 -123
README.md CHANGED
@@ -1,136 +1,235 @@
1
  ---
 
 
 
 
 
2
  dataset_info:
3
- - config_name: default
4
- features:
5
- - name: Age
6
- dtype: int64
7
- - name: Avg Monthly GB Download
8
- dtype: int64
9
- - name: Avg Monthly Long Distance Charges
10
- dtype: float64
11
- - name: Churn
12
- dtype: int64
13
- - name: Churn Category
14
- dtype: string
15
- - name: Churn Reason
16
- dtype: string
17
- - name: Churn Score
18
- dtype: int64
19
- - name: City
20
- dtype: string
21
- - name: CLTV
22
- dtype: int64
23
- - name: Contract
24
- dtype: string
25
- - name: Country
26
- dtype: string
27
- - name: Customer ID
28
- dtype: string
29
- - name: Customer Status
30
- dtype: string
31
- - name: Dependents
32
- dtype: int64
33
- - name: Device Protection Plan
34
- dtype: int64
35
- - name: Gender
36
- dtype: string
37
- - name: Internet Service
38
- dtype: int64
39
- - name: Internet Type
40
- dtype: string
41
- - name: Lat Long
42
- dtype: string
43
- - name: Latitude
44
- dtype: float64
45
- - name: Longitude
46
- dtype: float64
47
- - name: Married
48
- dtype: int64
49
- - name: Monthly Charge
50
- dtype: float64
51
- - name: Multiple Lines
52
- dtype: int64
53
- - name: Number of Dependents
54
- dtype: int64
55
- - name: Number of Referrals
56
- dtype: int64
57
- - name: Offer
58
- dtype: string
59
- - name: Online Backup
60
- dtype: int64
61
- - name: Online Security
62
- dtype: int64
63
- - name: Paperless Billing
64
- dtype: int64
65
- - name: Partner
66
- dtype: int64
67
- - name: Payment Method
68
- dtype: string
69
- - name: Phone Service
70
- dtype: int64
71
- - name: Population
72
- dtype: int64
73
- - name: Premium Tech Support
74
- dtype: int64
75
- - name: Quarter
76
- dtype: string
77
- - name: Referred a Friend
78
- dtype: int64
79
- - name: Satisfaction Score
80
- dtype: int64
81
- - name: Senior Citizen
82
- dtype: int64
83
- - name: State
84
- dtype: string
85
- - name: Streaming Movies
86
- dtype: int64
87
- - name: Streaming Music
88
- dtype: int64
89
- - name: Streaming TV
90
- dtype: int64
91
- - name: Tenure in Months
92
- dtype: int64
93
- - name: Total Charges
94
- dtype: float64
95
- - name: Total Extra Data Charges
96
- dtype: int64
97
- - name: Total Long Distance Charges
98
- dtype: float64
99
- - name: Total Refunds
100
- dtype: float64
101
- - name: Total Revenue
102
- dtype: float64
103
- - name: Under 30
104
- dtype: int64
105
- - name: Unlimited Data
106
- dtype: int64
107
- - name: Zip Code
108
- dtype: string
109
  ---
110
- ## Telco Customer Churn
111
 
112
- **This dataset is a valuable resource for exploring and predicting customer churn in the telecommunications industry. It provides a comprehensive snapshot of customer demographics, service usage patterns, billing information, and churn status, making it ideal for training machine learning models to predict customer churn and develop effective customer retention strategies.**
113
 
114
- **Content and Structure:**
115
 
116
- The dataset is structured in a tabular format, with each row representing a unique customer and each column containing attributes about that customer.
117
 
118
- * **Customer Demographics:** Features like gender, age, marital status, and dependents provide insights into customer profiles.
119
- * **Service Usage:** Details customer subscriptions to services such as phone, internet, multiple lines, online security, online backup, device protection, tech support, and streaming options.
120
- * **Billing Information:** Provides data on tenure, contract type, payment method, monthly charges, and total charges.
121
- * **Churn Information:** Includes labels indicating whether a customer churned, the reason for churn (if applicable), and churn scores for analysis.
122
 
123
- **Data Collection and Curation:**
124
 
125
- This dataset is a fictional dataset created by IBM data scientists as a sample dataset for exploring customer churn prediction. It is not based on real-world data and should be treated as a simulation for learning and experimentation.
126
 
127
- **Usage Examples:**
128
 
129
- * **Customer Churn Prediction:** Train classification models to predict churn based on customer demographics, service usage, and billing information.
130
- * **Customer Segmentation:** Analyze the dataset to identify customer segments with different churn probabilities, allowing for targeted retention strategies.
131
- * **Feature Engineering:** Experiment with feature engineering techniques to improve churn prediction model accuracy.
 
132
 
133
- **Additional Information:**
134
 
135
- * **Industry Relevance:** Relevant for businesses in the telecommunications industry and other sectors that deal with customer churn.
136
- * **Ethical Considerations:** This is a fictional dataset and does not contain real personal or sensitive information.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ language:
3
+ - en
4
+ pretty_name: Telco Customer Churn
5
+ size_categories:
6
+ - 1K<n<10K
7
  dataset_info:
8
+ - config_name: default
9
+ features:
10
+ - name: Age
11
+ dtype: int64
12
+ - name: Avg Monthly GB Download
13
+ dtype: int64
14
+ - name: Avg Monthly Long Distance Charges
15
+ dtype: float64
16
+ - name: Churn
17
+ dtype: int64
18
+ - name: Churn Category
19
+ dtype: string
20
+ - name: Churn Reason
21
+ dtype: string
22
+ - name: Churn Score
23
+ dtype: int64
24
+ - name: City
25
+ dtype: string
26
+ - name: CLTV
27
+ dtype: int64
28
+ - name: Contract
29
+ dtype: string
30
+ - name: Country
31
+ dtype: string
32
+ - name: Customer ID
33
+ dtype: string
34
+ - name: Customer Status
35
+ dtype: string
36
+ - name: Dependents
37
+ dtype: int64
38
+ - name: Device Protection Plan
39
+ dtype: int64
40
+ - name: Gender
41
+ dtype: string
42
+ - name: Internet Service
43
+ dtype: int64
44
+ - name: Internet Type
45
+ dtype: string
46
+ - name: Lat Long
47
+ dtype: string
48
+ - name: Latitude
49
+ dtype: float64
50
+ - name: Longitude
51
+ dtype: float64
52
+ - name: Married
53
+ dtype: int64
54
+ - name: Monthly Charge
55
+ dtype: float64
56
+ - name: Multiple Lines
57
+ dtype: int64
58
+ - name: Number of Dependents
59
+ dtype: int64
60
+ - name: Number of Referrals
61
+ dtype: int64
62
+ - name: Offer
63
+ dtype: string
64
+ - name: Online Backup
65
+ dtype: int64
66
+ - name: Online Security
67
+ dtype: int64
68
+ - name: Paperless Billing
69
+ dtype: int64
70
+ - name: Partner
71
+ dtype: int64
72
+ - name: Payment Method
73
+ dtype: string
74
+ - name: Phone Service
75
+ dtype: int64
76
+ - name: Population
77
+ dtype: int64
78
+ - name: Premium Tech Support
79
+ dtype: int64
80
+ - name: Quarter
81
+ dtype: string
82
+ - name: Referred a Friend
83
+ dtype: int64
84
+ - name: Satisfaction Score
85
+ dtype: int64
86
+ - name: Senior Citizen
87
+ dtype: int64
88
+ - name: State
89
+ dtype: string
90
+ - name: Streaming Movies
91
+ dtype: int64
92
+ - name: Streaming Music
93
+ dtype: int64
94
+ - name: Streaming TV
95
+ dtype: int64
96
+ - name: Tenure in Months
97
+ dtype: int64
98
+ - name: Total Charges
99
+ dtype: float64
100
+ - name: Total Extra Data Charges
101
+ dtype: int64
102
+ - name: Total Long Distance Charges
103
+ dtype: float64
104
+ - name: Total Refunds
105
+ dtype: float64
106
+ - name: Total Revenue
107
+ dtype: float64
108
+ - name: Under 30
109
+ dtype: int64
110
+ - name: Unlimited Data
111
+ dtype: int64
112
+ - name: Zip Code
113
+ dtype: string
114
  ---
115
+ # Dataset Card for Telco Customer Churn
116
 
117
+ This dataset contains information about customers of a fictional telecommunications company, including demographic information, services subscribed to, location details, and churn behavior. This merged dataset combines the information from the original Telco Customer Churn dataset with additional details.
118
 
119
+ ## Dataset Details
120
 
121
+ ### Dataset Description
122
 
123
+ This merged Telco Customer Churn dataset provides a comprehensive view of customer attributes, service usage, location data, and churn behavior. This expanded dataset is a valuable resource for understanding churn patterns, customer segmentation, and developing targeted marketing strategies.
 
 
 
124
 
125
+ ## Uses
126
 
127
+ ### Direct Use
128
 
129
+ This dataset can be used for various purposes, including:
130
 
131
+ - **Customer churn prediction:** Develop machine learning models to predict which customers are at risk of churning, leveraging the expanded features.
132
+ - **Customer segmentation:** Identify different customer segments based on demographics, service usage, location, and churn behavior.
133
+ - **Targeted marketing campaigns:** Develop targeted marketing campaigns to retain at-risk customers or attract new customers, tailoring campaigns based on the insights derived from the merged dataset.
134
+ - **Location-based analysis:** Analyze customer churn trends based on specific locations, cities, or zip codes, and identify potential regional differences.
135
 
136
+ ### Out-of-Scope Use
137
 
138
+ The dataset is not suitable for:
139
+
140
+ - **Real-time churn prediction:** The dataset lacks real-time data, making it inappropriate for immediate churn prediction.
141
+ - **Personal identification:** While the dataset contains customer information, it is anonymized and should not be used to identify individuals.
142
+
143
+ ## Dataset Structure
144
+
145
+ The dataset is structured as a CSV file with 49 columns, each representing a customer attribute. The columns include:
146
+
147
+ - **Age:** The customer's age in years.
148
+ - **Avg Monthly GB Download:** The customer's average monthly gigabyte download volume.
149
+ - **Avg Monthly Long Distance Charges:** The customer's average monthly long distance charges.
150
+ - **Churn Category:** A high-level category for the customer's reason for churning.
151
+ - **Churn Label:** Indicates whether the customer churned.
152
+ - **Churn Reason:** The customer's specific reason for leaving the company.
153
+ - **Churn Score:** A score from 0-100 indicating the likelihood of the customer churning.
154
+ - **Churn Value:** A numerical value representing whether the customer churned (1 for churned, 0 for not churned).
155
+ - **City:** The city of the customer's residence.
156
+ - **CLTV:** Customer Lifetime Value.
157
+ - **Contract:** The customer's contract type.
158
+ - **Country:** The country of the customer's residence.
159
+ - **Customer ID:** A unique identifier for each customer.
160
+ - **Customer Status:** The customer's status at the end of the quarter (Churned, Stayed, or Joined).
161
+ - **Dependents:** Whether the customer has dependents.
162
+ - **Device Protection Plan:** Whether the customer has a device protection plan.
163
+ - **Gender:** The customer's gender.
164
+ - **Internet Service:** Indicates whether the customer subscribes to internet service.
165
+ - **Internet Type:** The type of internet service provider.
166
+ - **Lat Long:** The combined latitude and longitude of the customer's residence.
167
+ - **Latitude:** The latitude of the customer's residence.
168
+ - **Longitude:** The longitude of the customer's residence.
169
+ - **Married:** Indicates if the customer is married.
170
+ - **Monthly Charge:** The customer's total monthly charge for all their services.
171
+ - **Multiple Lines:** Whether the customer has multiple phone lines.
172
+ - **Number of Dependents:** The number of dependents the customer has.
173
+ - **Number of Referrals:** The number of referrals made by the customer.
174
+ - **Offer:** The last marketing offer the customer accepted.
175
+ - **Online Backup:** Whether the customer has online backup service.
176
+ - **Online Security:** Whether the customer has online security service.
177
+ - **Paperless Billing:** Whether the customer has paperless billing.
178
+ - **Partner:** Whether the customer has a partner.
179
+ - **Payment Method:** The customer's payment method.
180
+ - **Phone Service:** Whether the customer has phone service.
181
+ - **Population:** The estimated population of the customer's zip code.
182
+ - **Premium Tech Support:** Whether the customer has premium tech support.
183
+ - **Quarter:** The fiscal quarter for the data.
184
+ - **Referred a Friend:** Indicates if the customer has referred a friend.
185
+ - **Satisfaction Score:** The customer's satisfaction rating.
186
+ - **Senior Citizen:** Whether the customer is a senior citizen.
187
+ - **State:** The state of the customer's residence.
188
+ - **Streaming Movies:** Whether the customer has streaming movies service.
189
+ - **Streaming Music:** Whether the customer has streaming music service.
190
+ - **Streaming TV:** Whether the customer has streaming TV service.
191
+ - **Tenure in Months:** The number of months the customer has been with the company.
192
+ - **Total Charges:** The customer's total charges.
193
+ - **Total Extra Data Charges:** The total charges for extra data downloads.
194
+ - **Total Long Distance Charges:** The total charges for long distance calls.
195
+ - **Total Refunds:** The total refunds received by the customer.
196
+ - **Total Revenue:** The total revenue generated by the customer.
197
+ - **Under 30:** Indicates if the customer is under 30 years old.
198
+ - **Unlimited Data:** Whether the customer has unlimited data.
199
+ - **Zip Code:** The zip code of the customer's residence.
200
+
201
+ ## Dataset Creation
202
+
203
+ ### Curation Rationale
204
+
205
+ This merged dataset was created to provide a more comprehensive and detailed analysis of customer churn behavior. Combining multiple sources of data allows for a richer understanding of factors influencing churn.
206
+
207
+ ### Source Data
208
+
209
+ #### Data Collection and Processing
210
+
211
+ The dataset is derived from the original Telco Customer Churn dataset and additional data sources. The specific data collection and processing methods are not disclosed.
212
+
213
+ ## Bias, Risks, and Limitations
214
+
215
+ ### Bias
216
+
217
+ The dataset may exhibit biases due to the simulated nature of the original Telco Customer Churn data. It is essential to consider that the dataset may not accurately reflect the demographics, service usage, or churn patterns of actual telecommunications companies.
218
+
219
+ ### Risks
220
+
221
+ Using the dataset for real-world decisions without proper validation and understanding of its limitations can lead to inaccurate predictions and potentially biased outcomes.
222
+
223
+ ### Limitations
224
+
225
+ - **Simulated Data:** The dataset is based on simulated data and may not fully represent real-world customer behavior.
226
+ - **Limited Context:** The dataset may lack specific contextual information such as customer feedback or reasons for churn.
227
+ - **Potential Bias:** The simulated data may not fully capture the nuances of customer behavior and churn patterns, especially when combined with additional data sources.
228
+
229
+ ### Recommendations
230
+
231
+ Users should be aware of the dataset's limitations and potential biases. Consider the following:
232
+
233
+ - **Validation:** Validate the dataset's results against real-world data before making critical decisions.
234
+ - **Contextualization:** Include additional contextual information if available to improve model accuracy and insights.
235
+ - **Transparency:** Be transparent about the dataset's limitations and potential biases when communicating results.