Audience & Dates
Embedded Files
Objectives
Target: This competition is open to students of all levels residing in Cameroon. Participants may form teams of 2 to 4 members. Diversity in gender and field of study within teams is highly encouraged.
Deadline
Team registration closes on: 15th March 2025 at 23:59 PM GMT +1. here (registered teams will be contacted after deadline)
Project will be pitched at the Conference & Code on a github page will be sent along side project documentation report.
Full competition details and dataset can be found on the here
Project submission closes on: 25th March 2025 at 23:59 GMT+1 (only registered groups will be contacted for this)
Objective: We are looking to build a comprehensive interactive dashboard, fully implemented in Python, for the visualization and analysis of blood donation campaign data. The dashboard should showcase the richness of the dataset and provide valuable insights for optimizing blood donation campaigns. Participants will need to create a dashboard that addresses key questions for campaign organizers, helping them make data-driven decisions to improve the success of future blood donation campaigns.
Objectif : Nous cherchons à concevoir un tableau de bord complet et interactif, entièrement implémenté en Python, pour la visualisation et l'analyse des données des campagnes de don de sang. Ce tableau de bord devra mettre en valeur la richesse du jeu de données et fournir des informations précieuses afin d’optimiser les campagnes de don de sang. Les participants devront créer un tableau de bord répondant aux principales questions des organisateurs de campagnes, les aidant ainsi à prendre des décisions basées sur les données pour améliorer le succès des futures campagnes de don de sang.
Finalist:
QG ANALYTICS - TechSpectra - MK dir Winners - M-I-Flow - CBDH2025
TensorPulse - The Outsider - Equipe_Alpha - HEMOBOARD - Re:Tech
HOPE2 - CodeStorm - CrimsonUnity - SENDA CONSULTING - HOPE
Data Storytellers - Optimus - BloodHub - NK STAT CONSULTING
CodeFlow - ABO - Team Surfers - GoTech
Winners:
1st place: - NK STAT CONSULTING
2nd Place: - MK dir Winners
3rd Place: - The Outsider
Key Features to be Implemented
The dashboard should provide answers to the following questions, leveraging the provided dataset:
Map Donor Distribution:
Visualize the geographical distribution of blood donors based on their residential area (i.e., "Arrondissement de résidence" and "Quartier de Résidence").
Use a map to plot the locations of donors, highlighting regions with high or low participation.
Health Conditions & Eligibility:
Visualize the impact of health conditions (e.g., hypertension, HIV, asthma, diabetes) on blood donation eligibility.
Create charts or graphs that display the number of eligible vs. non-eligible donors based on these conditions.
Profiling Ideal Donors:
Use clustering techniques to group donors into similar profiles based on demographic and health-related features (e.g., age, gender, profession, health conditions).
Generate insights into the characteristics of the ideal blood donor.
Campaign Effectiveness:
Analyze past campaigns by examining the donation date and other demographic factors.
Visualize trends, such as the time of year when blood donations are highest or which demographics contribute more to campaigns.
Identify patterns in donor behavior over time.
Donor Retention:
Investigate donor retention by analyzing how often individuals return to donate blood.
Use demographic data to determine which factors (e.g., age, profession, region) correlate with repeat donations.
Survey/Feedback Sentiment Analysis:
If feedback text data is available (e.g., in the "Si autres raison préciser" column), perform sentiment analysis on the textual feedback provided by donors.
Classify the feedback into positive, negative, or neutral categories and visualize sentiment trends over time or by demographic group.
Blood Donation Eligibility Prediction Model (API):
As an additional challenge, participants are asked to build a machine learning model that predicts the eligibility of new donors based on demographic and health data.
This model should be wrapped in an API to allow easy integration into the dashboard for real-time predictions.
Expected Deliverables:
Expected Deliverables:
A Fully Functional Python Dashboard:
The dashboard should be interactive, displaying all the visualizations and insights mentioned above.
Candidates should use Python libraries such as Dash, Streamlit, or Plotly to create the dashboard interface.
The dashboard should support interactivity, allowing users to filter and drill down into the data (e.g., by region, age, health condition, etc.).
Codebase:
Provide all code for the dashboard, including data cleaning, visualization, and modeling.
The code should be well-commented and modular.
Machine Learning Model:
The machine learning model for predicting blood donation eligibility should be packaged as an API.
The API should accept inputs (e.g., age, health condition, profession) and return a prediction of whether the individual is eligible to donate blood.
Documentation:
A comprehensive README file explaining the functionality of the dashboard, the tools used, and any assumptions made during development.
Clear instructions on how to run the dashboard and interact with the visualizations.
If applicable, explain how to use the prediction model API.
Judging Criteria:
Judging Criteria:
The submissions will be evaluated based on the following criteria:
Functionality:
Does the dashboard meet all the required features and functionality as outlined in the competition description?
Are the visualizations interactive and easy to understand?
Usability:
Is the dashboard user-friendly and intuitive? Can users easily filter data and navigate through the insights?
Is the user interface clean and professional?
Data Insights:
How well does the dashboard showcase the richness of the dataset? Are the insights meaningful and actionable for blood donation campaigns?
Are the visualizations clear and easy to interpret?
Innovation:
Does the submission go beyond the basic requirements, adding creative features or advanced visualizations?
If the bonus feature is implemented, how well does the machine learning model integrate with the dashboard?
Code Quality:
Is the code well-organized, modular, and properly commented?
Is the code reusable and maintainable?
Presentation Judges
Dr. Wagou Irène
Page updated
Google Sites
Report abuse