Business Analytics Project: Exploring Business Data for Improvements

This business analytics project involves a comprehensive analysis of a provided dataset to improve business processes. The project begins with an understanding of the dataset, identifying both nominal and numerical variables, and descriptive statistics. It proceeds to discover relationships among attributes, applying normalization techniques such as 1NF, 2NF and 3NF to understand functional dependencies. The project then delves into potential business analysis tasks, employing data mining techniques like regression analysis on key attributes. The regression model generated, with R-squared and adjusted R-squared values of 40%, allows for conclusions on how increases in independent variables impact the dependent variable ($sale amount), thereby aiding in better decision-making and organizational improvements. The analysis includes references to the provided resources, demonstrating a structured approach to the problem.

BUSINESS ANALYTICS
Introduction
Business analytics help most organizations or companies in better understanding and improving
the structure and sales of companies. This may be achieved through conducting a thorough
business analysis using the dataset provided in excel. For efficient analysis to be conducted, we
will consider the following;
 We should be able to understand the dataset provided in excel
 Relationships discovered among variables
 Potential analysis for business suggestions and the reason why we target this
potential analysis and the potential benefits that the expected results might aid
in business improvement.
Task 1: Understanding the dataset
There are many nominal and numerical variables available in this dataset. Some of the nominal
variables include; Identifies the type of dwelling involved in the sale, Identifies the general
zoning classification of the sale and many more. Some of the numerical variables include; $sale
amount, month sold, linear feet of street connected to property and so on. From the excel output
we can observe that descriptive statistics has been performed and it includes the mean, standard
error, minimum and maximum of linear feet of street connected to property which are 10516.83,
261.216, 1300 and 215245 respectively. More details are presented in the excel output for this
attribute and other attributes, in excel there is a scatter plot between month sold (x-axis) and $
sale amount (y-axis) and from the graph we can see that the highest $sale amount is
approximately $750,000 which corresponds to month sold of around 2 (Nicholas, 2006).
Task 2: Relationships discovery among attributes

Here we attempt to discover the relationships that exist between some of these variables and we
apply normalization technique. Using the first normal form (1NF), we see that there are no
attributes that contain multiple values hence we can proceed to the second normal form (2NF).
Here the variable Identifies of data instances (ID) is the candidate key and again in this data there
are no composite primary key but the attribute $sale amount (sale price) is part of the primary
key. In the third normal form (3NF), the functional dependency holds if and only if at least one
of the following conditions is satisfied;
 One of the attribute is a super key
 One of the attribute is a prime variable
Therefore from the excel output we observe that the type of sale determines condition of sale and
condition of sale determines $sale amount hence transitive functional dependency.
Task 3: Potential business analysis tasks
Here we discuss data mining techniques such as regression on some of the attributes given in
dataset. From the output in excel we observe that both R-square and adjusted R-square are 40%
which indicates that the below model is good for fit for the selected attributes (Bremer, 2012).
Y ($sale amount) =61373.39+0.898*lot size in square feet+112.62*type 1
finished square feet+62.03*type 2 finished square feet+100.93*unfinished square feet of
basement area.
Conclusion
From the model above we can conclude that increase in the independent attributes leads to
increase in the dependent variable ($ sale amount) hence this help in improving business process
for better decision making and also brings benefits to the organizations.

References
Bremer, M. (2012). [online] Mezeylab.cb.bscb.cornell.edu. Available at:
http://mezeylab.cb.bscb.cornell.edu/labmembers/documents/supplement%205%20-
%20multiple%20regression.pdf [Accessed 22 May 2019].
Nicholas, J. (2006). [online] Sydney.edu.au. Available at:
https://sydney.edu.au/stuserv/documents/maths_learning_centre/descstats2010web.pdf
[Accessed 22 May 2019].