Study

This requires selecting an appropriate application, (i.e., being able to explain or predict a phenomenon), preparing the data, analyzing the data through visualization, creating a model, and reporting your results. You are required to use IBM Watson. After reporting the results you will a write summary and conclusion of your findings. That is,

1. Demonstrate the application of CRISP approach: understanding the problem, understanding the data, etc.

2. Analyzing the data through visualization using IBM Watson Analytics

3. Summary of your findings.

4. Conclusion.

The datasets are provided below. Grading criteria for the project is presented below.

 

    Levels of Achievement

 

Criteria

Below Expectations

Proficient

Exemplary

 

Problem Identification and decision support requirement   management 

Weight 20.00%

50 %

70 %

100 %

 

Data collection and management 

Weight 20.00%

50 %

70 %

100 %

 

Discovery/Analytic: Mastery of Analysis – Model selection and   perform analysis 

Weight 20.00%

50 %

70 %

100 %

 

Mastery of Visualization 

Weight 20.00%

50 %

70 %

100 %

 

Reporting/Displaying 

Weight 20.00%

50 %

70 %

100 %

 

Possible data sets to choose from: 

Possible data sets to choose from:

Data from Hackathon such as  the Lord of the Machines – Data Science Hackathon, DataHack Premier League, Mckinsey Analytics Online Hackathon,  etc.   (Be brave  to take challenging  problem …)

Possible sources are:

a. https://archive.ics.uci.edu/ml/index.php

b. https://www.kaggle.com/datasets

c. https://www.kdnuggets.com/datasets/index.html (Datasets for Data Mining and Data Science)

d.  open government dataset @ https://catalog.data.gov/dataset?groups=local

e. Dataset from URL: https://www.ibm.com/communities/analytics/watson-analytics-blog/guide-to-sample-datasets/

OR

f. Other datasets after getting approval from the instructors.

For example at URL: https://www.ibm.com/communities/analytics/watson-analytics-blog/guide-to-sample-datasets/

Marketing

WA_Fn UseC_ Marketing Campaign Eff UseC_ FastF.csv
Quickly analyze test market campaigns based on responses, revenue and other key metrics. Predict who will respond to which campaign by which channel and why. Increase the likelihood of responses and quality of leads in future campaigns.

WA_Fn UseC_ Marketing Campaign Plan_ GroceryS.csv
Using this dataset around Coupons, you can quickly analyze marketing campaigns based on responses, revenue and other key metrics. Predict who will respond to which campaign, which channel they will use and why and thereby increase the likelihood of responses and quality of leads in future campaigns.

WA_Fn UseC_ Marketing Customer Value Analysis.csv
Understand customer demographics and buying behavior. Use predictive analytics to analyze the most profitable customers and how they interact. Take targeted actions to increase profitable customer response, retention and growth.

Finance

WA_Fn UseC_ Accounts Receivable.csv
Understand the factors of successful collection efforts. You can Predict which customers will pay fastest and recover more money and improve collections efficiency.

WA_Fn UseC_ Banking Loss Events 2007-14.xlsx
Understand hidden patterns and trends by combining various fields in Explore. Predict the leading drivers of a target, for example, using ‘Recovery Amount’ as the target.

Customer Support

WA_Fn UseC_ Telco Customer Churn.csv
Predict behavior to retain customers. You can analyze all relevant customer data and develop focused customer retention programs.

Operations

WA_Fn UseC_ Operations Dem Planning_ BikeShare.csv
Understand product and service usage characteristics based on operational and external information. Predict the likely placement or time a service will be used to optimize revenue. Finally, increase the likelihood of positive sentiments by having the right product available at the right time.