INDIAN SCHOOL BOUSHER

(Affiliated to the Central Board of Secondary Education, New Delhi)

(Affiliation No: 6630203 and School No: 90252)
P.O. Box: 2886, Postal Code: 130, Al Awabi, Al Ansab, Sultanate of Oman
ACADEMIC YEAR 2024 - 2025
REVISION WORKSHEET
Subject: Artificial Intelligence (843) Grade: XII Sections: A, B
Name of the Student: LF: Ms. Aurea & Ms. Sehba

PART B – UNIT 1
CAPSTONE PROJECT
S.NO Questions Marks
1. The train test split is a technique for evaluating the performance of a 01
machine learning algorithm. Which machine learning algorithm can it be
used?
a) Regression
b) Clustering
c) Classification
d) Deep Learning
Ans: Classification.
2. Mean Square Error (MSE) is the most commonly used regression loss 01
function. Identify one feature of MSE.
a) It is sensitive to outliers.
b) It is used on data, conditioned on the output variables.
c) It is good to use if the target data is normally distributed around a
median value.
d) It should be compared with Mean Absolute Error, where the optional
prediction is the mean.
Ans: It is sensitive to outliers.
3. Adding a non-important feature to a linear regression model may result in 01
___________.
a) Increase in R-square
b) Decrease in R-square
c) No change
d) Exponential Change
Ans: Increase in R-square
4. First four steps of writing a Python code to find out RMSE values of the 01
model are given. Arrange them in proper order.
a) Splitting the data into training and test.
b) Reading the data
c) Fitting simple linear regression to the training set.
d) Import required libraries
i) b–d–a–c
ii) d–c–b–a
iii) a–b–c–d
iv) d–b–a–c
Ans: d – b – a – c
5. An optimum AI model should have a __________ value less than 180. 01
a) Mean Square Error.
b) Mean Absolute Error
1 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
 c) Quantile Loss
d) Root Mean Square
Ans: Root Mean Square Error
6. What is K-NN algorithm and how does it determine the category for a new 02
instance.
Ans: K-Nearest Neighbour algorithm is one of the simplest Supervised
Learning-based Machine Learning algorithms. The K-NN algorithm assumes
similarity between the new case and the existing cases and assigns the new
instance to the category that matches the existing cases the most closely.
7. Can MSE be negative value? Why? Give the equation to calculate MSE? 02
Ans: MSE cannot be a negative value. The difference between the predicted
and actual values can be negative. However, these differences are squared.
Hence, all results are either positive or zero.

8. Define the 5 stages of Design Thinking. 02

Ans: Empathise – conduct research to better understand your users.
Define – Define the challenges, use your research to observe users’ current
problems.
Ideate – Brainstorm to arrive at various creative solutions.
Prototype – Build representation of one or more ideas.
Test – Test you models and gain user feedback.
9. Differentiate between training set and test set 02
Ans: The training dataset is used to fine-tune the machine learning model
and train the algorithm.
The test dataset is a set of data used to impartially evaluate how well the
final model fits the training dataset.
10. Explain the term Time Series Decomposition. 02
Ans: Time series decomposition is a fundamental step in time series
analysis because it helps in understanding the different contributing factors
in the data and can aid in making forecasts or predictions. It considers a
series as a combination of level, trend, seasonality and noise components.
11. How are MSE and RMSE related? What is their range? Are they sensitive to 04
outliers?
Ans: One of the most used regression loss functions is MSE. We determine
the error in Mean-Squared-Error, also known as L2 loss, by squaring the
difference between the predicted and actual values and average it
throughout the dataset.
RMSE value of 0 implies that the model is perfectly fitted. The model and its
predictions perform better when the RMSE is low. A greater RMSE indicates
a substantial discrepancy between the residual and the ground truth. The
RMSE of a good model should be less than 180.
12. Jain has a small dataset consisting of only 550 rows. Which training and 04
testing technique should he use? Give reasons.
Ans: For Jain's small dataset with only 550 rows, a simple train-test split is
a feasible approach.
Simplicity and Efficiency: A train-test split divides the data into two parts,
typically 80% for training and 20% for testing, which is easy to implement
and less computationally demanding. Given the dataset size, this simplicity
can be an advantage.
2 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
 Quick Evaluation: A train-test split provides a straightforward way to quickly
assess model performance, which can be beneficial for smaller datasets.
Avoiding Overfitting with Small Data: By keeping the model evaluation
separate from training, Jain can reduce overfitting while still testing the
model on a reasonable amount of data.
14. What is MAPE? How it can be used to analyse large datasets? Explain. 04
Ans: The accuracy of a forecasting technique is determined by the Mean
Absolute Percentage Error (MAPE). It represents the average of the
absolute percentage errors of each entry in a dataset. Large data sets may
typically be effectively analysed using MAPE, which requires that dataset
values should be other than zero.
MAPE = (1/n) * ∑ (| actual value – predicted value | / |actual value|) *100
Additionally, when using absolute percentage errors, the problem of
positive and negative errors cancelling each other out is eliminated.
15. Explain the two Data Modelling Approach. 04
Ans: Descriptive model – “describe” or summarise raw data and turn it into
something that can be understood by people. They are historical analytic
models. They enable us to understand how past behaviours may affect
future outcomes. Example: - company’s total inventory.
Predictive model – The ability to “predict” what might happen is the
foundation of predictive analysis. Based on data, predictive analysis gives
businesses actionable insights. Example: - creation of credit score.
16. What are the reasons for choosing Train Test Split Evaluation? 04
Ans: The reasons for choosing this technique are it can be used in large
dataset, to estimate the machine learning model’s performance on new
data that was not used to train the model, better computational efficiency
and a quick overview of model performance.
UNIT 2: Model Lifecycle
17. Which stage of AI model lifecylce is referred by the statement “Collecting 01
data from various reliable and authentic sources”?
a) Data Acquisition
b) Modelling
c) Data Exploration
d) Problem Scoping
Ans: Data Acquisition
18. The initial insights gained help to get an understanding of the data and 01
later on, help in algorithm selection, metrics choice, etc. This complete
procedure is called __________.
a) Exploratory Data Analysis
b) Data Exploration
c) Problem Scoping
d) Design Phase
Ans: Exploratory Data Analysis
19. The __________ is a subset of data employed for unbiased evaluation of 01
model fit and hyper parameter tuning.
a) Validation dataset
b) Training dataset
c) Testing dataset
d) Raw dataset
Ans: Validation dataset

3 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
20. Which of the following is the first stage of an AI Model life Cycle? 01
a) Build
b) Design
c) Testing
d) Scoping
Ans: Build
21. Two sources of authentic data are _________ and ________. 01
a) Government Websites, cameras or sensors
b) Real world data
c) Human entries
d) Corrupted data
Ans: Government websites and cameras
22. Describe the purpose of the following steps of the AI model cycle: 02
a. Data Exploration
b. Modelling
Ans: Data Exploration – After gathering data, the processes such as data
cleaning to locate missing values, eliminating worthless data, performing
basic statistical analysis are carried out.
Modelling – It is the process through which several models based on
graphical data can be constructed and even tested for advantages and
disadvantages.
23. Why is design phase an iterative process? 02
Ans: The design process is considered an iterative process because it
involves a cyclical and repetitive approach to problem-solving and
refinement. Rather than being a linear sequence of steps, design is an
ongoing loop where each iteration brings about improvements based on
feedback, testing, and evaluation.
24. What is the purpose of Data Exploration? 02
Ans: After gathering data, the processes such as data cleaning to locate
missing values, eliminating worthless data, performing basic statistical
analysis such as drawing graphs and comparing different properties of the
data set are carried out. It is useful to see which elements are more
essential and what the overall trend of the data is.
25. What is training data? When is testing of the model conducted? 02
Ans: Train Dataset: Used to fit the machine learning model.
Test Dataset: Used to evaluate the fit machine learning model.
26. What helps narrow down the number of feasible solutions in the modelling 02
phase?
Ans: Problem definition, i.e., understanding the problem. Data gathering,
Feature definition, AI model construction, and Evaluation helps in narrow
down the number of feasible solutions.
27. You are planning to create an AI model for the health-care industry. You 04
create a feasibility report. What information according to you should be
included in the feasibility report?
Ans: Identify the specific healthcare problem the AI model aims to solve,
Describe the types of data required, Specify the types of AI and machine
learning algorithms that could be used, explain how the model will
integrate into current healthcare systems and practices, Estimate the costs
involved in data collection, model development, computational resources,
and implementation, Summarize the feasibility of developing the AI model
based on the information collected.
4 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
28. Explain AI project cycle and project scoping in detail. 04
Ans: AI project lifecycle encompasses three main stages:
a) Project scoping b) Design or Build phase c) Deployment in production.
Project scoping:
i) The first fundamental step when starting an AI initiative is scoping and
selecting the relevant use cases that the AI model will be built to
address.
ii) This stage involves the planning and motivational aspects of your
project.
iii) Garbage in, garbage out: This means if the data you collect is not
good, you won’t be able to build an effective AI algorithm, and your
whole project will collapse.
iv) It is crucial to define the strategic business objectives and desired
outcomes of the project, select all the different stakeholder’s
expectations, anticipate the key resources and steps, and define the
success metrics.
29. Explain the terms over-fitting, under-fitting, and perfect-fit in terms of 04
model testing.
Ans: Overfitting – A model is overfitted when it is trained with a lot of data.
Underfitting – It usually happens when we have less data to build an
accurate model.
Perfect fit – This happens when our model is able to predict with good
accuracy on both testing and validation data.
30. What is Data Acquisition? List few authentic sources of data. 04
Ans: For data analysis, one must collect data from credible sources. Real
world data can be strange and deceptive. Human entries are always prone
to mistakes. Hence, the data needs to be relevant and authentic. Authentic
data can be gathered from government websites, devices such as camera
and sensors, purchases, transactions, registrations, other public surveys
and records.
31. A separate validation dataset is used for evaluation during training to 04
monitor how well model generalises, avoiding bias and overfitting. What are
the other few things considered during this stage?
Ans: a) The volume of test data can be huge, that provides data
complexities.
b) Human biases might have a negative impact on the testing phase, thus,
data validation is critical.
c) As the system may deal with the sensitive data, regulatory compliance
and security testing are essential.
d) Due to the sheer volume of data, performance testing is crucial.
e) If the AI solution requires data from other systems, systems integration
testing is crucial.
UNIT 3: Story Telling Through Data
32. Which of the following is NOT a key element of a data story? 01
a) Immersive Experience
b) Data
c) Visuals
d) Narrative
Ans: Immersive experience.
33. _________ are a popular type of financial chart which is used to represent 01
the price movements of an asset.
5 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
 a) Bar charts
b) Line charts
c) Candle charts
d) Maps
Ans: Candle charts
34. _________ refers to data that consists of numerical values or 01
measurements.
a) Numeric data
b) Text
c) Charts
d) Facet Grids
Ans: Numeric data
35. ____________ is the first step involved in Data story. 01
a) Creating a narrative
b) Keeping your audience interested.
c) Recognising the audience
d) Highlighting important information.
Ans: Recognising the audience
36. Data story telling is a ______ approach for communicating insights drawn 01
from data.
a) Iterative
b) Sequential
c) Procedural
d) Structured
Ans: Structured
37. What is the role of data, graphics and narrative in data storytelling? 02
Ans: When the proper graphics and narrative are combined with the correct
data, you have a data story that has the potential to impact and drive
change.
38. List the steps involved in telling an effective data story. 02
Ans: i) Recognising the audience
ii) Choosing the appropriate data and visualisations
iii) Highlighting important information.
iv) Creating a narrative.
v) Keeping your audience interested.
39. Why data storytelling is important? Give any two points to support your 02
answer.
Ans: Data storytelling is a powerful tool that can be used to communicate
complex ideas in a way that is easy to understand.
a) It aids in the dissemination of information/results.
b) It makes data memorable and simpler to remember in the long run.
40. What are geographic maps? 02
Ans: These maps contain geographic data which is the information that has
been plotted around a sphere. We utilise maps to give clarity and meaning
to our analysis when we have data relevant to certain regions and areas.
41. Write any four factors that make storytelling a powerful tool. 02
Ans: 1) It makes information more compelling
2) It helps with learning because stories are easy to remember.
3) It works for all types of listeners.
4) It allows interaction between the storyteller and the listener.
42. Why has data storytelling acquired a place of importance? 04
6 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025
 Ans: a) It is an effective tool to transmit human experience.
b) No matter how impressive an analysis, or quality the data is, it is not
going to compel change unless the people involved understand what is
explained through a story.
c) Stories that incorporate data and analytics are more convincing than
those based entirely on anecdotes or personal experience.
d) It helps to standardise communications and spread results.
e) It makes information memorable and easier to retain in the long run.
43. Explain how the three elements of a data story can influence change. 04
Ans: Data storytelling uses a structured approach to delivering data insights
that always includes a combination of three main elements: data, graphics,
and narrative. When a narrative is backed by data, it helps to explain to the
audience what is happening in the data and why a specific insight was
developed. When visuals are applied to data, they can enlighten the
audience to insights that they would not have noticed otherwise, such as
charts or graphs.
44. Write any five best practices for data storytelling. 04
Ans: 1) Always label your axes and give your plot a title.
2) When legends are required, use them.
3) Colours that are lighter on eye and in proportion should be used.
4) Avoid adding superfluous details to your visualisation.
5) When encoding time series, never use points for visualisation.
45. Briefly explain the purpose of data storytelling. 04
Ans: a) It is an effective tool to transmit human experience.
b) No matter how impressive an analysis, or quality the data is, it is not
going to compel change unless the people involved understand what is
explained through a story.
c) Stories that incorporate data and analytics are more convincing than
those based entirely on anecdotes or personal experience.
d) It helps to standardise communications and spread results.
e) It makes information memorable and easier to retain in the long run.
46. List down the steps involved in developing models from the data and telling 04
stories within them.
Ans: Data Exploration – a data analyst uses visual exploration to
understand what is in a dataset and the characteristics of the data, rather
than through traditional data management systems.
Feature Visualising – method for answering inquiries about what a model is
seeking.
Model creation – when we get to the stage of creating the model, we
typically realise that we need to comprehend how our data is being fitted to
the AI model.
Model comparisons – compare certain algorithms and techniques by looking
at their decision boundaries.

********

7 | Page
INDIAN SCHOOL BOUSHER/GRADE XI-XII/DEPARTMENT OF COMPUTER SCIENCE/2024-
2025