Introduction

Welcome to the DTP Data Science and Data Analysis Hackathon! This is your space to learn, build, and connect with Rwanda’s digital innovators.

 

Choose One of the following topics:

 

1. Climate Change and Environmental Sustainability

❖ Challenge: Develop predictive models to forecast local climate risks (e.g., floods,

droughts) using historical weather and environmental data.

❖ Methodology: Data cleaning, exploratory data analysis, feature engineering, time

series forecasting or machine learning models.

❖ Impact: Help communities prepare for climate events, enabling proactive

disaster management.

❖ Scalability: Models can be adapted for different regions worldwide.

(Inspired by Microsoft’s AI for Earth and Data Science Global Impact Challenge)[1]

 

2. Healthcare Predictive Analytics

❖ Challenge: Predict patient readmission risks or disease outbreaks from hospital

or public health datasets.

❖ Methodology: Data preprocessing, classification models, validation techniques,

and interpretability analysis.

❖ Impact: Improve healthcare resource allocation and patient care quality.

❖ Scalability: Applicable to various healthcare systems globally.

(Common in hackathons focusing on social good and healthcare)[1][2]

 

3. Financial Fraud Detection and Risk Assessment

❖ Challenge: Build models to detect fraudulent transactions or assess credit risk

using financial datasets.

❖ Methodology: Anomaly detection, supervised learning, feature importance

analysis, and model deployment strategies.

❖ Impact: Enhance security and trust in financial services.

❖ Scalability: Can be integrated into banking and fintech platforms worldwide.(Aligned with Data Open and FinTech hackathon themes)[1][3]

4. Supply Chain Optimization and Logistics

❖ Challenge: Optimize delivery routes or inventory management using real-time

and historical logistics data.

❖ Methodology: Data integration, optimization algorithms, predictive analytics,

and visualization dashboards.

❖ Impact: Reduce costs, improve delivery times, and increase sustainability.

❖ Scalability: Solutions can be scaled across industries and geographies.

(Inspired by hackathons focusing on logistics and business optimization)[1][3][4]



5. Social Good: Nonprofit Impact Measurement

❖ Challenge: Analyze nonprofit program data to measure impact and optimize

resource allocation.

❖ Methodology: Data aggregation, statistical analysis, visualization, and

storytelling through data.

❖ Impact: Help nonprofits improve effectiveness and transparency.

❖ Scalability: Frameworks can be adapted for various social causes globally.

(Based on Hack for Good and social impact hackathons)[1]

 

6. Bonus: Sales Forecasting for E-commerce Business

Scenario: You are consulting an e-commerce platform that sells electronic gadgets

globally. The CEO wants to understand sales trends and prepare for next quarter.

❖ Methodology: Focus on technical implementation and modeling approach:

Data Import & Cleaning: Import data from multiple CSVs using Python or R, Merge:

Orders, Products, Customers, Shipping, Handle missing values and inconsistencies, make

exploratory Data Analysis (EDA)

Forecasting Model Development: Apply moving averages, exponential smoothing,

regression, validate and compare model performance

Interactive Dashboard Creation: Build with Tableau / R Shiny / Julia include filters and

visuals for business stakeholders

❖ Impact: Demonstrate how the solution adds value or solves a real-world

problem:

Problem Solving: Regional Sales Decline

Diagnose reasons for lower sales in specific regions, propose and evaluate three

improvement solutions

Communication Task (Board Meeting): Present results in business terms and

communicate insights to executive-level stakeholders

❖ Scalability: Focus on the ability to generalize, scale, or adapt the solution:

Discussion on dashboard adaptability (embedded in dashboard task)

Potential to extend to new product categories or regions

Extend forecasting models to new seasons or real-time input.

 

For each challenge, learners should be required to:

Describe their methodology: data cleaning, modeling approach, validation, and tools

used.

Demonstrate impact: how their solution addresses the problem and benefits

stakeholders.

Tools and Technologies

    • Python, Jupyter, R

- [Top Data Science Hackathons in 2025](Link)