Skip to content

Data Analysis and Business Modeling Lab

CIS · Final · Summer 26

Section A · Batch 232

This browser can’t show the PDF here

Open it, and your phone’s PDF viewer takes it, or download it.

Read the questions as text

Read off the PDF by AI, so it may have mistakes. The PDF is the original.

Daffodil International University Department of Computing and Information Systems Lab Final: Summer-2026 Program: B.Sc. in CIS Course Code: BI334L Course Title: Data Analysis and Business Modeling Lab Task: Maternal Health Risk Assessment Total Marks: 40 Submission Type: Individual Submission Format: Sharing Jupyter Notebook in Kaggle (Handle: MAK) and pdf file on AI Professor Exam Scenario You are given the Maternal Health Risk Factors Dataset [1] collected from pregnant patients at General Hospital, Kurigram, Bangladesh. Build a reproducible data-analysis workflow that explores maternal health indicators and develops a model for predicting the binary Risk Level (Low/High). Dataset Variables Age, Systolic BP, Diastolic BP, Blood Sugar (BS), Body Temperature, BMI, Previous Complications, Pre-existing Diabetes, Gestational Diabetes, Mental Health, Heart Rate, and Risk Level. Instructions - Use the provided maternal-health dataset only; do not substitute another dataset. - Write all code in Python. The notebook must run from top to bottom without errors. - For every analysis, show the relevant output and provide a brief interpretation in your own words. - Use a fixed random_state where applicable so results are reproducible. - Do not copy numerical findings from the accompanying dataset article; compute results from the supplied data. - Clearly justify preprocessing and modeling choices. Avoid unnecessary preprocessing. Marking Rubric Component | Marks | Assessment focus Data understanding & quality | 6 | Inspection, cleaning decisions, observations Exploratory data analysis | 8 | Relevant visualizations and interpretation Statistical analysis | 8 | Correct tests, hypotheses, computation, interpretation Preprocessing & modeling | 12 | Sound pipeline, three classifiers, metrics, comparison Interpretation & conclusion | 6 | Risk-factor evidence, critical discussion, conclusion Total | 40 | Task 1 - Data Understanding and Quality Assessment [6 marks] - Load the dataset and report its shape, column names, data types, and target variable. (2 marks) - Check missing values and duplicate rows. Report what you find and explain any action taken. - Produce appropriate descriptive statistics and state two initial observations. (2 marks) Task 2 - Exploratory Data Analysis [8 marks] - Visualize the distribution of Risk Level and comment on class balance. (2 marks) - Create suitable plots to examine BMI and Blood Sugar across Risk Level. Interpret the patterns. (2 marks) - Create a scatter plot for Age versus BMI, distinguishing observations by Risk Level. Interpret it. (2 marks) - Create a correlation heatmap for appropriate numeric variables and identify two notable relationships. Task 3 - Statistical Analysis [8 marks] - Construct a contingency table for Pre-existing Diabetes versus Risk Level and perform a Chi-Square test of independence. State H0/H1 and interpret the p-value at alpha = 0.05. (4 marks) - Compare BMI between patients with and without Pre-existing Diabetes using an appropriate two-sample test. State H0/H1, report the test statistic and p-value, and interpret the result. (4 marks) Task 4 - Preprocessing and Predictive Modeling [12 marks] - Prepare X and y. Apply only necessary encoding/scaling and justify the steps. Use a stratified train/test split. (3 marks) - Train any three suitable classifiers. At least one must be a linear model and at least one a tree-based model. (4 marks) - Evaluate all models using Accuracy, Precision, Recall, F1-score, and a Confusion Matrix. Present a comparison table. (3 marks) - Select the best model and justify your choice, giving particular attention to correctly identifying high-risk cases. (2 marks) Task 5 - Interpretation and Viva Readiness [6 marks] - Identify the three most influential or informative risk factors using an appropriate evidence-based method (e.g., coefficients, feature importance, or permutation importance). (3 marks) - Write a concise conclusion covering the main findings, one limitation, and one practical implication for maternal-health risk assessment. (3 marks) Academic Integrity and Submission Notes - This is an individual examination. Identical or substantially copied notebooks may be treated as plagiarism. - You must be able to explain your code, statistical tests, preprocessing choices, model results, and conclusions during the viva. - Results shown without executable code or meaningful interpretation may receive reduced marks. References [1] Mojumdar, M. U., Sarker, D., Assaduzzaman, M., Shifa, H. A., Sajeeb, M. A. H., Islam, O., Bari, M. S., Alam, M. J., & Chakraborty, N. R. (2025). Maternal health risk factors dataset: Clinical parameters and insights from rural Bangladesh. Data in Brief, 59, 111363. https://doi.org/10.1016/j.dib.2025.111363