Machine Learning Business Profession Quiz Answer

Machine Learning for Business Professionals course is offered by “Coursera”. In this post you will get Machine Learning Business Profession Quiz Answer | 100% Correct Answers.

Machine Learning for Business Professionals

Offered By ”Google Cloud”

⭐ 4.6 stars (3305 ratings)

Instructor: Google Cloud Training

N.B. We attempted our best to keep this site refreshed for our clients for nothing. You can likewise contribute by refreshing new inquiries or existing inquiry answer(s). There are numerous inquiries on our site, it is difficult for us to check them consistently. It will be extraordinary on the off chance that you can assist us with updating the site. Just let us know if you find any new questions through mail or comment . We will attempt to refresh the inquiry/answer ASAP.

Use “Ctrl+F” To Find Any Questions Answer. & For Mobile User You Just Need To Click On Three dots In Your Browser & You Will Get A “Find” Option There. Use These Option to Get Any Random Questions Answer.

All the Qwiklabs Answer Of week- 1 Are Given In The Below Video

Week- 1

Use Cases

1. For each of the following 5 scenarios, determine the difficulty or specificity of each suggested use case.

Scenario 1: Discrete manufacturing e.g. Caterpillar
Say that you are a discrete manufacturing company and you are looking to monitor and track the health of your shipping fleet. You have 50 large container ships that deliver goods.

Use Case: Predict the optimal number of months between ship maintenance cycles to maintain good performance

Too Unspecific – we dont know the number of container ships

Too Specific – each ship is likely to have a different tracking device that gives different data about Lat/Long + speed
Too Easy – this is already a solved problem and the company should buy a commercial off the shelf solution
Too Difficult – predicting ship maintenance cannot be done

Just Right – building a proof-of-concept model, given enough high quality data, could be possible. Let’s try it!

2. Scenario 1: Discrete manufacturing e.g. Caterpillar
Say that you are a discrete manufacturing company and you are looking to monitor and track the health of your shipping fleet. You have 50 large container ships that deliver goods.

Use Case: Use data from on-board ship IoT devices to track device power consumption and predict cost savings

Too Unspecific – we don’t know what we’re measuring
Too Specific – our insights are not generalizable enough across all ships

Too Easy – why bother, this is a solved problem
Too Difficult – getting sensors on ships is too difficult
Just Right – putting sensors on ships to track speed and consumption could get us good data for modeling purposes.

3. Scenario 1: Discrete manufacturing e.g. Caterpillar
Say that you are a discrete manufacturing company and you are looking to monitor and track the health of your shipping fleet. You have 50 large container ships that deliver goods.

Use Case: Predict the total amount of revenue to be generated by a ship’s cargo once sold

Too Unspecific – we dont know what we’re calculating
Too Specific – the insights wouldn’t be useful
Too Easy – you wouldnt need a Machine Learning model to calculate revenue from goods sold (it would just be units * price)

Too Difficult – you can’t predict the revenue

Just Right

4. Scenario 1: Discrete manufacturing e.g. Caterpillar
Say that you are a discrete manufacturing company and you are looking to monitor and track the health of your shipping fleet. You have 50 large container ships that deliver goods.

Use Case: Predict whether or not to be in the shipping business

Too Unspecific – this question is too vague for Machine Learning. What are we actually trying to predict?

Too Specific
Too Easy – this has been solved before many times
Too Difficult – you can never predict where a company should be in a certain businesses or not (regardless of how good the questions you ask are)

Just Right

Module 2 Quiz

1. A model trained to predict life expectancy is trained on data collected from Norwegians. This model is successful in that market. However, when it is used in nearby Lithuania, performance suffers considerably. This is most likely an issue of:

Cleanliness

Coverage
Completeness

2. The success of expert systems demonstrates that:

Small sets of if-then statements are too brittle to work as models in most domains
It’s already possible to create explainable ML models for many domains

3. Moore’s Law describes

The rate at which the density, and therefore, the processing power, of computer chips doubles

The maximum amount of speedup we can expect when increasing the number of processors used to perform a task
The relationship between the size of computer chips and the amount of electric power they need to consume

4. The reliability of Moore’s Law meant that for a long time, one could expect the time it took to compute a workload to halve every two years. Even though Moore’s Law is no longer reliable, recent advances in this area or these areas have led to a similar sort of optimism.

ASICs
Cloud Computing

Storage technology
Distributed Computing

All the Qwiklabs Answer Of week- (2-4) Are Given In The Below Video

Week- 2

Module 3 Quiz

1. A reasonable percentage of the data to reserve for evaluation is:

20%
100%

2. A survey is administered over the internet via links in ads that appear on the front page of popular local newspapers. The survey asks sensitive questions about health and medical history. This could be an example of:

Automation bias
Selection bias

Confirmation bias
Reporting bias

3. Fill in the blanks with the appropriate fairness criteria. ________________ leads to models that use the same threshold, regardless of sub-group whereas _____________ leads to models that use different thresholds but which lead to similar distributions of predictions.

Group unaware, Demographic parity
Group unaware, Equal opportunity

Demographic parity, Equal opportunity

Week- 3

Module 4 Quiz

1. Unstructured data are

Datasets that aren’t well organized or which use inconsistent conventions

Data that can’t be easily or objectively compared

2. True or False: It can be a good idea to decompose large use cases into smaller ones, and align each smaller use case to its own model.

True
False

Week- 4

Module 5 Quiz

1. Traditional data storage solutions suffer from which of the following:

All of the above
Poor performance

Inseparable combination of compute and storage
Siloed access policies and technologies
Lack of centralized metadata stores

2. Data enrichment refers to the practice of:

Cleaning data by removing errors and inconsistencies

Using analytics and machine learning to compute values no originally present in the data but which have business value
None of these
Joining data with third party data