Common Problem Types

Question 1

What are the six common problem types that analysts work with mentioned in the course?

Answer 1

1. Making predictions
2. Categorizing things
3. Spotting something unusual
4. Identifying themes
5. Discovering connections
6. Finding patterns

Question 2

What does making predictions entail for an analyst?

Answer 2

It involves data from measurements of variables to determine the likelihood of an event happening.

e.g.
1. A hospital monitoring system using a remote patient monitoring system + their patient data to predict and prevent future health problems in chronically ill patients.
2. Marketing analytics- predicting the best marketing strategy.

Question 3

What do analysts do when they need to categorize things?

Answer 3

They assign information/ subjects into different groups/clusters based on common features shared in each group.

1. Categorizing employees based on their performance in different activities.
2. Classifying customer service calls based on certain key words to correlate with customer satisfaction scores.

Question 4

What problem reveals itself when analysts uncover something unusual?

Answer 4

Data that is different from the norm i.e. signifivantly higher or lower than average.

1. Surge in sales or surge in registered students
2. Smartwatches that set off alarms when health data doesn’t trend normally.

Question 5

What do analysts do when they need to identify themes in their analysis?

Answer 5

They need to summarize information/data categories/ groups into broader overarching concepts.

1. Analysts summarizing employee categories into a productivity measure.
2. User Experience analysts, examining user interaction data.

Question 6

When do analysts need to discover connections?

Answer 6

When they are examining similar/same challenges faced by different entities to combine data and insights inorder to co-operatively address them.

1. A third party logistics company woking with another company to get determine appropriate schedule changes to increase number of on time deliveries.

Question 7

What does a data analyst need to do to find patterns?

Asking “has it happened before” - an important question in the initial phases.

Answer 7

They need to analyze hisotrical data to see of an event/occurence has occured repetitively across certain intervals or conditions in-order to predict when it may happen again.

1. In E-Commerce, analysts looking through transaction data to understand customer buying habits at certain times of the year like holiday trends.
2. Understanding when equipment has failed before and if there were any conditions that preceded that failure.