Page 165 - Artificial Intellegence_v2.0_Class_12
P. 165
B. Fill in the blanks.
1. Human biases in selecting test data can negatively impact .
2. is an example of detrimental effects that can occur over time on the AI model.
3. For a balanced dataset, may be a useful choice for evaluating a classification model.
4. The is used to assist us in identifying the important factors associated with the problem.
5. The initial insights gained help to get an understanding of the data and later on, help in algorithm selection, metrics
choice, etc. This complete procedure is called .
6. The constitutes the largest proportion of the original dataset and is employed to train the machine
learning models.
7. The is a subset of data employed for unbiased evaluation of model fit and hyperparameter tuning.
C State hether the ollo ing statements are true or alse
1. If the data you collect is good, you will be able to build an effective AI algorithm, and your whole
project will collapse.
2. Human entries are always prone to mistakes.
3. Building the algorithm corresponds to Evaluation stage of the AI model cycle.
4. The AI model gets trained on the data fed to it and then is able to design a model which is
adaptive to the change in data.
5. Various AI development platforms provide substantial documentation to assist development teams.
D Match the ollo ing
1. Training set a. 4Ws Problem Canvas
2. BigML b. data cleaning
3. Data Exploration c. AI development Platform
4. Problem Scoping d. iterative process
5. Design phase e. fit the model
S CTI N B (Subjective Type Questions)
A Short ans er type uestions
1. What is the purpose of Data Exploration?
Ans. After gathering data, processes such as
• data cleaning to locate missing values
• eliminating worthless data (erroneous samples and outliers)
• performing basic statistical analysis such as drawing graphs (or any other visual representation) and comparing
different properties of the data set, are carried out. The initial insights gained help to get an understanding of
the data and later on, help in algorithm selection, metrics choice, etc. It is useful to see which elements are more
essential and what the overall trend of the data is.
2. During which stage of the AI model cycle should we take care of AI Ethics? What are the challenges to this process?
Ans. During the development of the AI model (modelling phase), care should be taken that the programmer is including
data, instructions, etc. that provides protection of all liberties for all citizens as per the fundamental rights of the
country. There are two main challenges—One is getting access to high-quality and standardised datasets and second
is being able to find skilled programmers who can develop reliable and high-quality machines.
M ode l L i f e cycl e 163

