Module 2: AI and digital transformation

Knowledge check: Introduction to AI decision-making models

Practice Assignment

1. What is the primary advantage of using decision trees?

  • They can only handle numerical data.
  • They are free from overfitting.
  • They require large amounts of data to work.
  • They are easy to interpret and visualize. ✅

Explanation:
 Decision trees are widely used because they are simple to understand and can be visualized easily, making them a popular choice for interpretable machine learning.

2. True or False: A random forest combines multiple decision trees, reducing overfitting and improving accuracy.

  • False
  • True ✅

Explanation:
A random forest reduces overfitting and improves accuracy by combining multiple decision trees through techniques like bagging and averaging.

3. What is the primary role of the hidden layers in a neural network?

  • To make the final prediction.
  • To collect raw data for training.
  • To store the output of the neural network.
  • To combine and extract patterns from input features. ✅

Explanation:
Hidden layers in a neural network process input data, combining and extracting patterns to create meaningful representations that are passed to the next layer.

4. Which of the following statements about neural networks is true? Select all that apply.

  • Weights are initialized by the activation function.
  • The activation function is used to decide which information passes through a neuron. ✅
  • The Sigmoid activation function produces outputs between 0 and 1, making it suitable for probability-based tasks. ✅
  • In a neural network, the updating of the weights takes place only during the training phase. ✅

Explanation:
Activation functions control the flow of information, and the sigmoid function is especially useful for probabilities. Weights are updated only during training as part of the optimization process.

5. Which of the following scenarios best describes how a neural network adjusts its weights during training?

  • Weights are fixed during training to avoid instability.
  • The output layer directly updates all weights in the network.
  • The network adjusts weights randomly to improve accuracy.
  • Weights are updated based on feedback from the error rate or prediction error. ✅

Explanation:
 Neural networks adjust weights during training using optimization techniques like gradient descent, guided by the error rate, to minimize prediction errors.

Knowledge check: Digital transformation

Practice Assignment

6. In the context of digital transformation, organizations increasingly rely on AI to enhance operations. Which of the following best describes AI’s contribution to this transformation?

  • Eliminating the need for data in decision-making
  • Automating processes to improve efficiency ✅
  • Replacing all human roles in an organization
  • Reducing the importance of change management

Explanation:
AI enhances operations during digital transformation by automating repetitive tasks, enabling faster and more efficient workflows, and improving decision-making with data-driven insights.

7. True or False: An online bookstore struggling with clunky checkouts, delivery delays, and a lack of personalized recommendations can succeed by focusing on the four domains of digital transformation.

  • True ✅
  • False

Explanation:
By focusing on the four domains of digital transformation—customer experience, operational processes, business models, and workforce empowerment—the bookstore can resolve issues like clunky checkouts, delivery delays, and lack of personalization.

8. Which technologies are commonly used to drive digital transformation? Select all that apply.

  • Internet of Things (IoT) ✅
  • Cloud Computing ✅
  • Artificial Intelligence (AI) ✅
  • Legacy Software

Explanation:
Technologies such as IoT, Cloud Computing, and AI are foundational to driving digital transformation by enabling better connectivity, scalability, and intelligent decision-making. Legacy software is typically a barrier, not a driver, to digital transformation.

9. A retail company wants to optimize inventory management using AI. Which approach is most effective?

  • Migrate data to an on-premises server.
  • Implement IoT devices for customer feedback collection.
  • Use predictive analytics to forecast demand. ✅
  • Develop an AI chatbot to improve customer service.

Explanation:
Predictive analytics helps optimize inventory management by forecasting demand accurately, reducing overstock and stockouts, and improving supply chain efficiency.

10. An online bookstore is struggling to compete due to poor customer experiences, delivery delays, and a lack of personalized recommendations. Which strategy best incorporates the four domains of digital transformation to address these issues?

  • Offer flat discounts to increase sales and collaborate with coffee shops for hybrid events.
  • Use customer behavior data to personalize recommendations, implement automated inventory systems, and train employees on digital tools. ✅
  • Focus solely on redesigning the website and adding a wish list feature.
  • Outsource delivery services and add more staff for manual inventory tracking.

Explanation:
This strategy leverages digital tools to enhance customer experience, streamline operations, and empower employees, addressing all four domains of digital transformation effectively.

Knowledge check: Business challenges

Practice Assignment

11. A technology company wants to scale its artificial intelligence (AI) model to manage increasing user demand and large datasets. During this process, it encounters system efficiency and data security challenges. What is the most critical issue the company needs to address?

  • Ensuring data privacy and security at scale. ✅
  • Reducing the size of training datasets to save resources.
  • Designing algorithms that prioritize efficiency over scalability.
  • Avoiding the use of distributed computing to keep costs low.

Explanation:
Scaling AI involves handling larger datasets and more user interactions, making data privacy and security critical challenges. Companies must ensure compliance with privacy laws and protect user data.

12. A healthcare provider adopts an AI tool to predict risks for critical conditions. Doctors express concern about how the AI reaches conclusions, specifically when recommending treatments. Why is it critical for the AI model to provide interpretable results?

  • To make AI models faster by simplifying algorithms.
  • To completely eliminate the need for human oversight in treatment planning.
  • To allow stakeholders to understand and trust the AI’s decision-making process. ✅
  • To ensure users can easily modify the AI model’s decisions directly.

Explanation:
 Interpretability builds trust by helping doctors and stakeholders understand how AI reaches its conclusions, which is essential in sensitive fields like healthcare.

13. A financial institution uses an AI model to predict the risk of loan defaults based on customer data, including credit scores, income, and spending patterns. A customer is denied a loan and asks why the AI made this decision. How should the institution provide clarity on the AI’s decision?

  • Avoid using interpretability techniques and simply provide the customer with the raw decision.
  • Simplify the AI model by removing features that are hard to interpret.
  • Use Local Interpretable Model-agnostic Explanations (LIME) to highlight the features that contributed most to the decision.
  • Use SHapley Additive exPlanations (SHAP) to calculate and explain each feature’s contribution to the decision. ✅

Explanation:
SHAP provides a clear and detailed breakdown of how each feature impacts the AI model’s decision, helping explain the reasoning to customers and stakeholders effectively.

14. What is one key challenge businesses face when integrating AI solutions with existing Customer Relationship Management (CRM) systems?

  • Eliminating all manual processes from the CRM system.
  • Using AI only to automate routine customer service tasks.
  • Replacing the CRM system entirely with the AI solution.
  • Ensuring seamless data flow between AI and the CRM system. ✅

Explanation:
One of the biggest challenges is integrating AI with existing systems to ensure smooth data exchange, which is necessary for AI to enhance CRM functionality.

15. A company uses AI for targeted advertising campaigns based on user data. What steps should it prioritize to comply with GDPR and ensure ethical practices?

  • Avoiding all forms of data storage to comply with GDPR.
  • Ensuring AI algorithms remain confidential and avoid transparency with regulators.
  • Using AI to bypass user consent requirements.
  • Ensuring that users provide informed consent for data collection and usage. ✅

Explanation:
To comply with GDPR and maintain ethical standards, companies must prioritize user consent and transparency in how data is collected, stored, and used.

Module quiz: AI and digital transformation

Practice Assignment

16. True or False: Change management is not one of the key domains of digital transformation, which focuses on helping employees adapt to new tools and processes.

  • False ✅
  • True

Explanation:
Change management is a crucial domain of digital transformation, focusing on helping employees adapt to new tools and processes.

17. True or False: A company introducing AI-based chatbots to improve customer support is an example of digital transformation.

  • False
  • True ✅

Explanation:
Introducing AI-based chatbots to improve customer support is an example of digital transformation, as it involves adopting new technologies to enhance business operations.

18. Which of the following correctly describes the purpose of weights in a neural network?

  • Weights store the input data for the neural network.
  • Weights are irrelevant to a neural network and are used in decision trees only.
  • Weights determine the connections between neurons and influence how data flows through the network. ✅
  • Weights are the output of the neural network.

Explanation:
Weights play a critical role in determining the strength of connections between neurons and how input data is transformed in a neural network.

19. A retail company wants to adopt digital transformation technologies to improve its operations and customer experience. It is considering using AI for personalized recommendations, IoT devices for real-time inventory tracking, and cloud platforms for data management. Which benefits are these technologies likely to provide?

  • Automated physical delivery of products to customers.
  • Improved customer engagement through personalized recommendations. ✅
  • Real-time monitoring and optimization of inventory. ✅
  • Scalable data storage and enhanced collaboration. ✅

Explanation:
AI, IoT, and cloud platforms collectively enable personalized customer experiences, efficient inventory management, and scalable data operations.

20. Netflix leverages digital transformation by using data analytics and machine learning to enhance user experience. Which of the following best describes how Netflix applies these technologies?

  • Monitoring user devices for potential technical issues.
  • Ensuring that the subscriber cannot leave the subscription to Netflix.
  • Providing personalized content recommendations based on viewing history. ✅
  • Automatically generating movies and TV shows without the need for human involvement.

Explanation:
Netflix uses data analytics and machine learning to analyze viewing patterns and provide personalized recommendations to enhance user experience.

21. True or False: A neural network is created to mimic how computer chips work on a deeper level and some biological concepts.

  • True
  • False ✅

Explanation:
A neural network is inspired by biological neural networks but is not designed to mimic computer chips. It models how data is processed through interconnected neurons.

22. True or False: Machine Learning (ML) is a subset of Artificial Intelligence (AI) that focuses on enabling systems to learn and improve from data.

  • True ✅
  • False

Explanation:
Machine Learning (ML) is a subset of AI, focusing on systems that learn and improve from data without explicit programming.

23. Suppose you have a list of students with their study hours and attendance. You are using this list to create a decision tree that predicts whether a student will Pass or Fail based on how many hours they have studied and their attendance. Now, if you use the following dataset to train a decision tree to classify "Pass" or "Fail" based on study hours and attendance:

All other paths should result in a Fail. The decision tree predicts the result for a new student who studied for 6 hours and attended 80% of the classes. What is the most likely output if study hours is evaluated first followed by attendance?

  • Fail
  • The decision tree becomes stuck and cannot proceed to a leaf node because there is insufficient data to complete the decision flow.
  • Pass ✅
  • Cannot determine

Explanation:
The new student studied for 6 hours (greater than 4) and had 80% attendance (greater than 70%). Based on the decision tree’s rules, this results in “Pass.”

24. What is the role of neurons in a neural network?

  • Neurons store data for future analysis.
  • Neurons determine the structure of the entire network.
  • Neurons provide feedback to the user.
  • Neurons process inputs and pass the output to the next layer. ✅

Explanation:
Neurons in a neural network receive inputs, process them using weights and activation functions, and pass the output to the next layer.

25. A Random Forest model is trained to predict house prices based on features like size, location, and number of bedrooms. The predictions from individual trees for a specific house are:

Tree 1: $200

Tree 2: $220

Tree 3: $210

What is the final prediction of the Random Forest model for the house?

  • $200
  • $220
  • $250
  • $210 ✅

Explanation:
The final prediction of a Random Forest model is typically the average of predictions from individual trees:
(200+220+210)/3=210.(200 + 220 + 210) / 3 = 210.

Leave a Reply