Q1
A real estate agency in a metropolitan area
wants to develop a model to predict house prices accurately. Design an experiment
using the suitable data mining algorithm to assist the agency in building a
predictive model for house prices.
Instructions:
1. Select a suitable dataset containing information about
real estate properties, including
features like area, bedrooms, bathrooms, location, and sale prices.
2. Perform necessary data preprocessing steps, including
handling missing values, encoding categorical variables, and scaling numerical
features.
3. Choose
relevant features that could
influence house prices and build data mining model to predict prices based on
these features.
4. Choose
appropriate data mining task that
could influence house prices and build a model to predict prices based on these
features.
5. Split the dataset into training and testing sets and
find which Model Selection method gives best accuracy among the following
(a)Hold Out (b) K-Fold Cross Validation (c) Stratified K-Fold Cross
Validation
6.
Evaluate the
performance of the trained model using appropriate metrics such as Mean Absolute
Error (MAE), Mean Squared Error (MSE), R-squared (R2) and
accuracy
Analyze the coefficients of the of the model to interpret the impact of each feature on house prices.
8.
Predict house prices by reading multiple unknown values as a Data Frame.
9.
Plot the model with input data
values along with unknown values
1 Provide
recommendations to the real estate agency based on the analysis and
interpretation of the model results.
Q2
Educational institutions strive to support
student success and improve academic outcomes by identifying students who may
be struggling and providing them with appropriate interventions. Design an
experiment using the suitable clustering algorithm to categorize students based
on their academic performance, assisting in the identification of at-risk
students who may benefit from intervention and support programs.
Instructions:
1. Select a suitable dataset containing student academic records, including
features such as grades, attendance, study hours, participation in
extracurricular activities, and socio-economic background.
2. Perform necessary data preprocessing steps to prepare the data for
clustering analysis, ensuring data quality and consistency.
3. Apply the suitable clustering algorithm to the preprocessed data to
partition students into distinct clusters based on their similarities in
academic performance metrics.
4. Analyze the resulting clusters to understand the unique characteristics
and performance levels of students within each cluster.
5. Develop targeted intervention strategies for students in each performance
category, including academic support programs, mentoring, counseling, and
resources allocation tailored to the needs of students in each cluster.
6. Provide recommendations to the educational institution based on the
analysis and interpretation of the student performance categories to improve
academic outcomes and support student success.
Q3
Organizations strive to optimize their employee
hierarchy to ensure fair compensation, talent development, and organizational
effectiveness. Design an experiment using the suitable Clustering algorithm to
analyze the salary-based employee hierarchy within an organization, assisting
in identifying potential areas for restructuring or improvement to enhance
organizational performance.
Instructions:
1. Select
a suitable dataset containing information about employees within the
organization, including features such as employee ID, salary, department, job
title, years of experience, and performance ratings.
2. Perform
necessary data preprocessing steps to prepare the data for clustering analysis,
ensuring data quality and consistency.
3.
Apply
the suitable Clustering algorithm to the preprocessed data to identify
hierarchical structures based on employee salaries.
4. Analyze
the resulting clusters to understand the grouping of employees based on salary
levels and identify potential areas for optimization or restructuring.
5. Develop
recommendations for optimizing the employee hierarchy, including strategies for
salary adjustments, promotions, talent development, and succession planning,
based on the analysis of hierarchical clusters.
6. Provide
actionable insights to organizational stakeholders based on the analysis and
interpretation of the employee hierarchy optimization results to enhance
organizational performance and employee satisfaction.
Q4
Retailers aim to enhance customer satisfaction
and increase sales by delivering personalized shopping experiences tailored to
the preferences and needs of individual customers. Design an experiment using
the suitable clustering algorithm to segment customers based on spatial density,
their purchasing behavior and demographics, assisting in targeted marketing and
personalized customer engagement strategies.
Instructions:
1. Select
a suitable dataset containing customer transaction data from a retail store,
including features such as purchase history, frequency, recency, monetary
value, demographics, and location.
2. Perform
necessary data preprocessing steps to prepare the data for clustering analysis,
ensuring data quality and consistency.
3. Apply
the suitable algorithm to the preprocessed customer data to identify clusters
of customers based on their spatial density in the feature space.
4. Analyze
the resulting clusters to understand the distinct segments of customers based
on their purchasing behavior and demographics.
5. Develop
targeted marketing strategies for each customer segment, including personalized
promotions, product recommendations, and communication channels tailored to the
preferences and needs of customers in each cluster.
6. Provide
recommendations for implementing and operationalizing the segmented customer
strategy to enhance customer engagement and increase sales.
Q5
A telecommunications
company is experiencing high customer churn rates and wants to develop a
predictive model to identify customers at risk of churning. Design an
experiment using appropriate data mining algorithm to assist the company in
building a model for predicting customer churn.
Instructions:
1.
Select
a suitable dataset containing information about telecommunications customers,
including features such as account length, international plan, voicemail plan,
number of customer service calls, and churn status.
2.
Perform
necessary data preprocessing steps, including handling missing values, encoding
categorical variables, and scaling numerical features.
3.
Choose
relevant features that could influence customer churn and build a model to
predict churn status based on these features.
4. Split the dataset into training and testing sets and
find which Model Selection method gives best accuracy among the following
(a)Hold Out (b) K-Fold Cross
Validation (c) Stratified K-Fold Cross Validation
5.
Evaluate
the performance of the trained model using appropriate classification metrics
such as accuracy, precision, recall, and F1-score on the testing data.
6.
Analyze
the model structure by selecting the best splitting criterion to interpret the
key factors driving customer churn and identify actionable insights for the
telecommunications company.
7.
Predict the risk of churning by reading multiple unknown values as a Data
Frame.
8.
Provide
recommendations to the company based on the analysis and interpretation of the
model results.
9.
Outline
the steps the company should take to implement and utilize the predictive model
effectively in their operations.
Q6
Email spam continues to be a significant
problem, with potentially harmful consequences such as phishing attacks and
malware distribution. Design an experiment using probability based
classification algorithm to develop a model for detecting email spam, helping
users filter out unwanted and potentially dangerous emails from their inboxes.
Instructions:
1.
Select
a suitable dataset containing labeled emails, distinguishing between spam and
non-spam (ham) emails.
2.
Perform
necessary data preprocessing steps to convert the textual data into numerical
features suitable for classification.
3.
Build
a data mining model to classify emails as spam or non-spam based on the
presence or absence of certain words.
4.
Split
the dataset into training and testing sets and train the model using the
training data.
5.
Evaluate
the performance of the trained model using classification metrics such as
accuracy, precision, recall, and F1-score on the testing data.
6.
Analyze
the model's predictions and misclassifications to understand its effectiveness
in distinguishing between spam and non-spam emails.
7.
Predict
email spam by reading multiple unknown values as a Data Frame
8.
Provide
recommendations for users based on the analysis and interpretation of the model
results to improve email security and reduce the risk of falling victim to
email scams or cyber attacks.
Q7
Retail stores strive to maximize sales and
enhance customer satisfaction by understanding purchasing patterns and
optimizing product offerings. Design an experiment using the different
appropriate data mining algorithm to perform market basket analysis for a
retail store, assisting in identifying associations between products and
recommending strategies for improving sales and customer experience.
Instructions:
1.
Select
a suitable dataset containing transaction records from the retail store, where
each transaction lists the items purchased by a customer.
2.
Perform
necessary data preprocessing steps to prepare the transaction data for market
basket analysis, ensuring data quality and consistency.
3.
Apply
the suitable algorithm to the transaction data to generate frequent itemsets,
setting appropriate parameters such as minimum support threshold.
4.
Generate
association rules from the frequent itemsets, considering metrics such as
confidence, lift, and support to identify meaningful associations between
products.
5.
Analyze
the generated frequent itemsets and association rules to uncover patterns and
insights that can inform decisions related to product placement, promotions,
and cross-selling strategies by using different appropriate data mining
algorithms
6.
Provide
recommendations to the retail store based on the analysis and interpretation of
the market basket analysis results to optimize sales and enhance the customer
shopping experience.
