1st question:

U+ Bank has introduced a new credit card for customers with annual income more than USD 100000. Given the scenario, which is the most useful rule type?

Practice exam says correct answer is: Adaptive Model

I think correct answer is: When rule or Decision table, which can be used in Proposition filter rule and strategy respectively

Predictive Model. Incorrect answer and so is Adaptive model

2nd question:

What is the difference between predictive and adaptive analytics?

Practice exam says correct answer is: Predictive models can predict a continuous value.

I think correct answer is: Predictive models predict customer behavior.

Adaptive models use customer data as predictors -  this is incorrect because it says customer data but customer properties are predictors

Predictive models have evidence - this is incorrect as Adaptive models have evidence

Thank you for the private message. Based on your input I realised the first question can be confusing, therefore I have changed it to this: 

U+ Bank has introduced a new credit card for which no historical customer behavior is known. U+ wants to offer this credit card on the customer personalized web portal. Given the scenario, which is the rule type you must use?

Can you please answer it now? The new question will be available in production later today.

Regarding the second question, please see my comments in bold to your comments:

>>>I think correct answer is: Predictive models predict customer behavior.

Both adaptive and predictive models predict customer behavior.

Adaptive models use customer data as predictors -  this is incorrect because it says customer data but customer properties are predictors

Both adaptive and predictive models use customer properties (customer data) as predictors. E.g. Age, House, Address, Houe Owner etc. 

>>> Predictive models have evidence - this is incorrect as Adaptive models have evidence

Predictive models are based on historical data. Adaptive models accumulate data on the fly. Both models are automatically created on a set of data, which we call evidence.

>>> Predictive models can predict a continuous value.

This is the correct answer. Adaptive models can only distinguish between binary outcomes. In the outcome definition, you differentiate between positive and negative behavior. You can not predict a continuous value, like for e.g. "the loan amount we should give to customers", "the minimum claim value which should automatically be approved" etc. Such continuous values can only be predicted using predictive models.