Business Analytics DEC 2026
December 2026 Examination
Q1 A regional retail chain has recently integrated several years’ worth of sales data from multiple stores into a central Excel dataset. During initial analysis, analysts discovered many missing values in the ‘delivery amount’ column due to inconsistent point-of-sale entry practices across store locations. Senior management is concerned about the impact of missing data on profitability reporting and trend analysis, especially as the company prepares for a major expansion into new markets. The analytics lead must recommend an approach for handling missing values that balances data integrity and analytical reliability while maintaining comparability across different store datasets.How should the analytics lead apply appropriate imputation techniques (mean, median, or mode) to address missing values in the ‘delivery amount’ column? Discuss the implications of each method on analytical outcomes and propose a step-by-step Excel-based workflow that ensures the integrity and usability of the final dataset for profitability analysis across all store locations. (10 Marks)
Ans 1.
Introduction
The absence of delivery numbers can degrade store profits and hide changes in the demand of customers. Yet, replacing every single blank by a common amount would result in a completely different issue. Stores could have a different clientele, charge different delivery fees, and operate under different price policies. Therefore, the analytics manager should analyze the reasons why transactions aren’t being recorded prior to deciding on an imputation technique. Mean, median and mode are able to provide useful estimates when applied to comparable transactions. The aim is to protect valuable information, without having to present estimates as recorded factual
Its Half solved only
Buy Complete assignment from us
Price – 190/ assignment
NMIMS Online University Complete SolvedAssignments session DEC 2026
Last date 28 Oct 2026
buy cheap assignment help online from us easily
we are here to help you with the best and cheap help
Contact No – 8791514139 (WhatsApp)
OR
Mail us- [email protected]
Our website – www.assignmentsupport.in
Q2 (A) A major airline is seeking to improve its on-time performance after receiving negative feedback from both customers and industry regulators. The airline’s analytics team has developed a probability-based machine learning model that predicts flight delays using historical flight data, weather patterns, aircraft types, and operational variables. However, industry experts argue that the model’s accuracy is limited by incomplete data and the assumption that all input variables are independent, which may not hold in a dynamic, multi-airport environment. The airline is considering whether to invest further in enhancing data quality or to revise its modelling approach to better reflect interdependencies.Evaluate the effectiveness of relying solely on probability-based models for predicting flight delays in the presence of incomplete data and variable interdependencies. What are the key risks and benefits for the airline, and how would you justify improvements to the modelling approach to maximise operational impact? (5 Marks)
Ans 2A.
Introduction
Airline companies require delays predictions that allow quick operational decision-making. Probability estimates could help you allocate the gates, staff, and standby aircraft. Incomplete records or false independence assumptions can render predictions misleading. The airline needs to enhance data quality and model design in conjunction, and evaluate the success by operational
Q2 (B) FinStat, a mid-sized financial advisory firm, uses SLR to predict loan default rates from customer income. In a quarterly report, the analytics manager notes a low SSR compared to TSS, resulting in a high R-squared near 0.9. However, the regression diagnostic review reveals heavy-tailed residuals not following the normal distribution, and some evidence of autocorrelation, especially during economic downturns. Leadership is considering using this model for major risk management decisions but is concerned about validity under stressed conditions.Critically evaluate the reliability of FinStat’s regression model for strategic loan portfolio management, given the high R-squared but violations in normality and autocorrelation assumptions. Justify your conclusions by weighing the trade-offs between statistical goodness-of-fit and model assumption breaches, recommending improvements for robust risk assessment. (5 Marks)
Ans 2B
Introduction
FinStat’s explanatory accuracy is encouraging, but it cannot create reliable risk-based forecasts for loan risks for downturns. Autocorrelation and large tails indicate there is a lot of uncertainty. Management must evaluate predictions errors and model assumptions. You should evaluate a linear regression with stress prior to establishing lending policies or major limit on risk in the