D600 Task 3: Principal Component
Analysis Report
A. GitLab Repository
A subgroup and project were successfully created in GitLab for Task 3. The project was correctly
cloned into the IDE, and commits were made after the completion of each rubric-aligned
section. Each commit included a descriptive message and timestamp to ensure proper version
tracking and traceability. The GitLab repository URL was submitted in the “Comments to
Evaluator” section as instructed, and a full branch history was exported and included in the final
submission.
B1. Proposal of Question
The research question addressed in this project is: “How do structural characteristics such as
square footage, number of bedrooms, backyard space, age of home, renovation quality, and
previous sale price, along with neighborhood factors including crime rate, school rating, distance
to city center, employment rate, property tax rate, local amenities, and transportation access,
influence housing prices in the dataset?” This question reflects a real-world organizational
concern relevant to property valuation and real estate analytics. The problem is suited for linear
regression modeling because it examines the relationships between a continuous dependent
variable (home price) and multiple quantitative explanatory variables.
, B2. Defined Goal
The goal of the analysis was to build a predictive model of home prices using principal
component analysis (PCA) to reduce multicollinearity, followed by a linear regression model.
This goal supports informed decision-making in real estate by identifying which components of
housing data—such as location, size, or community amenities—have the greatest impact on
price. The analysis remains within the scope of the dataset and is appropriately supported by
the variables provided.
C1. PCA Use
Principal component analysis was used to transform a set of correlated continuous variables
into a smaller number of uncorrelated components. The PCA reduced redundancy in the data,
resolved issues of multicollinearity, and improved the reliability of the subsequent linear
regression model. The expected outcome of PCA was a reduced dataset where the retained
components captured the majority of the total variance, allowing for effective and interpretable
model training.
C2. PCA Assumption
One fundamental assumption of PCA is that the principal components are uncorrelated linear
combinations of the original variables. This assumption ensures that each component captures a
unique dimension of the dataset’s variance and contributes independently to the regression
model. In this analysis, the assumption held, as the PCA transformed the original correlated
variables into orthogonal components.
D1. Variable Identification
The continuous variables selected for PCA and regression included SquareFootage,
NumBedrooms, BackyardSpace, CrimeRate, SchoolRating, AgeOfHome, DistanceToCityCenter,
EmploymentRate, PropertyTaxRate, RenovationQuality, LocalAmenities, TransportAccess, and
PreviousSalePrice. These variables were logically aligned with the research question and were
confirmed to be continuous in nature. Non-continuous and problematic variables, such as
NumBathrooms, were excluded from the analysis based on data quality concerns.
D2. Standardized Data
Before performing PCA, all selected continuous explanatory variables were standardized using
Scikit-learn’s StandardScaler. This transformation ensured that each variable had a mean of zero
and a standard deviation of one, which is essential for PCA to give equal weight to all features
regardless of their original scale. The cleaned and standardized dataset was saved and
submitted as required.
Analysis Report
A. GitLab Repository
A subgroup and project were successfully created in GitLab for Task 3. The project was correctly
cloned into the IDE, and commits were made after the completion of each rubric-aligned
section. Each commit included a descriptive message and timestamp to ensure proper version
tracking and traceability. The GitLab repository URL was submitted in the “Comments to
Evaluator” section as instructed, and a full branch history was exported and included in the final
submission.
B1. Proposal of Question
The research question addressed in this project is: “How do structural characteristics such as
square footage, number of bedrooms, backyard space, age of home, renovation quality, and
previous sale price, along with neighborhood factors including crime rate, school rating, distance
to city center, employment rate, property tax rate, local amenities, and transportation access,
influence housing prices in the dataset?” This question reflects a real-world organizational
concern relevant to property valuation and real estate analytics. The problem is suited for linear
regression modeling because it examines the relationships between a continuous dependent
variable (home price) and multiple quantitative explanatory variables.
, B2. Defined Goal
The goal of the analysis was to build a predictive model of home prices using principal
component analysis (PCA) to reduce multicollinearity, followed by a linear regression model.
This goal supports informed decision-making in real estate by identifying which components of
housing data—such as location, size, or community amenities—have the greatest impact on
price. The analysis remains within the scope of the dataset and is appropriately supported by
the variables provided.
C1. PCA Use
Principal component analysis was used to transform a set of correlated continuous variables
into a smaller number of uncorrelated components. The PCA reduced redundancy in the data,
resolved issues of multicollinearity, and improved the reliability of the subsequent linear
regression model. The expected outcome of PCA was a reduced dataset where the retained
components captured the majority of the total variance, allowing for effective and interpretable
model training.
C2. PCA Assumption
One fundamental assumption of PCA is that the principal components are uncorrelated linear
combinations of the original variables. This assumption ensures that each component captures a
unique dimension of the dataset’s variance and contributes independently to the regression
model. In this analysis, the assumption held, as the PCA transformed the original correlated
variables into orthogonal components.
D1. Variable Identification
The continuous variables selected for PCA and regression included SquareFootage,
NumBedrooms, BackyardSpace, CrimeRate, SchoolRating, AgeOfHome, DistanceToCityCenter,
EmploymentRate, PropertyTaxRate, RenovationQuality, LocalAmenities, TransportAccess, and
PreviousSalePrice. These variables were logically aligned with the research question and were
confirmed to be continuous in nature. Non-continuous and problematic variables, such as
NumBathrooms, were excluded from the analysis based on data quality concerns.
D2. Standardized Data
Before performing PCA, all selected continuous explanatory variables were standardized using
Scikit-learn’s StandardScaler. This transformation ensured that each variable had a mean of zero
and a standard deviation of one, which is essential for PCA to give equal weight to all features
regardless of their original scale. The cleaned and standardized dataset was saved and
submitted as required.
