Description

Business Analytics

Apr 2026 Examination

 

 

Q1 An online food delivery marketplace collects vast arrays of structured (transaction times, payment amounts), semi-structured (order logs in JSON), and unstructured data (customer reviews and social media posts). The data science team faces difficulties integrating all these data types for insightful analytics, as trends in one type are often missed when isolated from others. Management now expects the team to use data type identification frameworks and integration strategies to unify the analysis and extract comprehensive business intelligence.Using the classification of data types discussed in the chapter, explain how you would apply these frameworks to integrate structured, unstructured, and semi-structured data for holistic analytics. What business benefits could arise from this integrated approach, and what challenges must you address in the preprocessing stage to enable unified insights? (10 Marks)

Ans 1.

Introduction

Food delivery platforms are very dynamic, and timely, fast, and complete data insights will be required to make decisions. These platforms produce huge amounts of organized records of transactions, semi-structured system logs, and unstructured discussions of customers in the form of reviews and social media posts. When such data forms are discussed separately, valuable trends are concealed and business knowledge is broken. To address this shortcoming, companies need to employ data classification models and integration models that enable the different data sources to collaborate with one another. The process of establishing a cohesive analytics will allow the management to have a more in-depth understanding of the customer behavior, the efficiency of organization, and market trends. Its Half solved only

Buy Complete assignment from us

Price – 190/  assignment

NMIMS Online University Complete SolvedAssignments  session DEC 2025

Last date 29 March 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. After implementing targeted process improvements based on customer survey analysis, Mehta E-commerce noticed variable results across different customer segments. While younger customers responded favorably to faster delivery, older demographics prioritized product quality and support. The analytics team utilized one-sample and two-sample hypothesis tests to quantify differences among these groups but struggled to interpret high p-values and overlapping confidence intervals. Senior management must decide whether to pursue uniform changes or segment- specific strategies in response to these findings, weighing the risk of misallocating resources and alienating certain customer groups.Assess the effectiveness of Mehta E- commerce’s application of hypothesis testing to support segment-specific versus uniform intervention strategies. How should management interpret high p-values and overlapping confidence intervals in this context, and what further analytical or sampling approaches could help justify a targeted customer satisfaction strategy? (10 Marks)

Ans 2.

Introduction

Hypothesis testing is an important part of the data-driven decision making where it is necessary to determine whether the difference in customer behavior or customer satisfaction is statistically significant, or it is merely the result of the random distribution. In the case of Mehta E-commerce, the application of one-sample and two-sample hypothesis tests to examine the responses given by the surveyed individuals according to their age as customer groupings is a structured effort by Mehta E-commerce to get out of the intuition and make strategic decisions based on findings. The issue, however, occurs when statistical outcomes like high p-values and overlapping confidence intervals are obtained, thus bringing doubt as to whether differences are strong. It is crucial to interpret these results properly because their incorrect interpretation may result in the implementation of uniform strategies which fail to

 

Q3(A). At EduGrowth Schools, the management team is analyzing the intricate connection between student absenteeism and academic achievement using a simple linear regression approach. They have compiled a robust dataset spanning several years, including precise attendance logs and academic scores for each term. However, the team observes that frequent absenteeism sometimes coincides with low performance, but they suspect external influences such as family background and health could also play a significant role. With stakeholders demanding targeted interventions, EduGrowth must produce a predictive framework that not only reveals this relationship but also guides future instructional support.Design a comprehensive regression-based framework that synthesizes the available absenteeism and academic data, accounting for potential external influences. Propose innovative strategies for model construction, validation, and practical intervention planning to maximize student academic outcomes while mitigating the effects of absenteeism. Justify each aspect of your design. (5 Marks)

Ans 3a.

Introduction

Understanding the relationship between student absenteeism and academic performance is essential for designing effective educational interventions. While simple linear regression can reveal general trends, real-world learning outcomes are influenced by multiple social and personal factors. For EduGrowth Schools, the challenge is not only to measure the impact of attendance on achievement but also to develop a predictive framework that supports targeted academic support. A well-structured regression-based approach can transform historical data

 

Q3 (B). A manufacturing conglomerate seeks to forecast production costs using multiple regression. The initial analysis includes variables such as raw material prices, labour costs, production volume, and maintenance hours; however, high-frequency fluctuations in these factors alongside new technologies being introduced have rendered existing models less predictive. Leadership wants a future-ready model that can anticipate volatility and evolving operational patterns while providing actionable insight for production planning.Develop a novel regression-based modelling and validation strategy that integrates external data sources, predictive scenario analysis, and adaptive model updating to future-proof cost forecasting. Outline how your strategy would balance immediate interpretability with long-term adaptability and support proactive, data-driven manufacturing decisions. (5 Marks)

Ans 3b.

Introduction

Forecasting production costs in a dynamic manufacturing environment requires models that can adapt to volatility and technological change. Traditional multiple regression models often struggle to remain accurate when operational conditions shift rapidly. For the manufacturing conglomerate, developing a future-ready cost forecasting framework is essential for maintaining profitability and operational efficiency. A modern regression-based strategy