This project analyzes over a century’s worth of mosquito vector data across Africa, with a focus on Anopheles species—key vectors for malaria. It aims to identify geographic, temporal, and environmental patterns, and apply machine learning to aid global health strategies for malaria control.
- Understand the distribution of Anopheles mosquito species across Africa.
- Reveal long-term sampling trends (1898–2016).
- Visualize geospatial spread using interactive maps.
- Predict mosquito presence using machine learning.
- Identify high-impact zones for health intervention and funding prioritization.
- Data Analysis:
pandas,numpy - Visualization:
matplotlib,seaborn,folium - Machine Learning:
scikit-learn(Random Forest Classifier, Permutation Importance) - Geospatial Mapping:
folium,MarkerCluster,HeatMap
- Displays dataset structure: column names, types, and missing values.
- Summarizes statistical distribution of values.
- Identifies key columns with missing data.
- Horizontal bar chart of the most frequently observed Anopheles species.
- Highlights An. gambiae as the most common malaria vector.
- Line graph illustrating sampling frequency from 1898 to 2016.
- Peak sampling activity occurs in the 2000s.
- Interactive Folium map showing sample locations across Africa.
- Color-coded markers representing vector species and countries.
- Model: Random Forest Classifier
- Top Predictors:
- Latitude
- Longitude
- Year of sampling
- Visualizations:
- ROC curve
- Confusion matrix
- Feature importance plots
- Predicts WHO malaria control intervention priorities.
- Analyzes feature importance for strategic planning.
- Generates a country-level heat map of priority scores.
- Recommends high-impact countries (e.g., Guinea Bissau, Equatorial Guinea) for urgent intervention.
- An. gambiae is the most sampled and significant vector.
- Temporal trends suggest increased surveillance from 2000 onward.
- Geospatial coverage is uneven; West & Central Africa are well-represented.
- Predictive power is strongest for geographic and temporal features.
- Data-driven models can inform national and global intervention strategies.
- Increase sampling in underrepresented regions.
- Prioritize recent data collection for timely decision-making.
- Use geographic and temporal trends to guide WHO resource allocations.
- Integrate vector mapping into national malaria control strategies.
Africa Vector Database (1898–2016)
Includes mosquito species data, geolocation, year of sampling, and associated metadata.
This project demonstrates the power of data science and machine learning in shaping public health strategies. With targeted interventions based on real-time analysis, health organizations can maximize the impact of limited resources in the fight against malaria.