Page 145 - Artificial Intellegence_v2.0_Class_12
P. 145
Stage 6: Data Preparation: his stage contains all the activities to build the dataset used in the subsequent
modeling stage. Activities to prepare data include
• data cleansing handling missing or invalid values, removing duplicates, applying correct formats),
• oining data from multiple sources files, tables, platforms), and
• the conversion of data to more useful variables.
Data preparation is usually the most time consuming procedure in a data science pro ect. Automating certain data
preparation steps in advance can speed up the process by minimising ad hoc preparation time.
. hat are hyperparameters hat is their purpose ive e amples of fe hyperparameters.
Ans. yperparameters are parameters hose values govern the learning process. hey are 'top level' parameters that
regulate the learning process and model parameters that come from it, as the prefi 'hyper' suggests. ince the
model cannot modify its values during learning training, hyperparameters are said to be e ternal to the model.
ome e amples of hyperparameters are
• he ratio of train test split
• Optimisation algorithms' learning rate e.g. gradient descent)
• he loss function that the model ill employ
• A neural net ork's number of hidden layers
• A clustering task's number of clusters
. plain the purpose of evaluation and deployment stage.
Ans. valuation tage
he data scientist
• utilises a testing set for predictive models that is separate from the training set but follo s the same
probability distribution and has a kno n outcome.) he testing set is used to assess the model and ad ust it as
necessary.
• or a final assessment, the final model is sometimes applied to a validation set as ell.
In addition, data scientists can use statistical significance tests to verify the model's accuracy.
Deployment tage
he model is deployed into the production environment or an equivalent test environment once it has been built
and authorised by the business sponsors. It is usually used in a restricted capacity until its effectiveness has been
thoroughly assessed. Deploying a model into a live business process frequently necessitates the involvement of
additional internal teams, skills, and technology.
ased/Application-b
ased questions:
C. Competency-based/Application-based questions:
ency-b
Compet
. onsider the follo ing statements containing an assertion and a reason
elect the appropriate option for the statements given above
a. oth A and are true and is the correct e planation of A
b. oth A and are true and is not the correct e planation of A
c. A is true but is false
d. A is alse but is true
i. Assertion (A): At the core of every AI model is ‘finding patterns in data.’
Reason (R): inding the right pattern is usually an iterative process.
Ans. b
ii. Assertion (A): onsider that the goal of an AI model is to predict an ans er such as "yes" or "no".
Reason (R): In such a case, predictive modeling can be used.
Ans. c
C apstone P roj e ct

