## Analysing and Evaluating Mobility in Northern Kenya
The project's aim is to identify the key factors influencing mobility in the North of Kenya. And specifically observe the role water scarcity plays in the region. With a multi-feature target, the analysis is done by applying unsupervised learning.
**BACKGROUND:** The ASAL region of Kenya has endured three severe droughts in the last decade (2010-2011, 2016-2017 and 2020-2022). The most recent drought (2020 - 2022) has also been the most severe and longest with widespread livelihood losses and massive displacement of populations.
**PROCESS:**
The dataset is based on in-situ surveys carried out in five counties in Northern Kenya in 2022-23 organized by United Nations International Organization for Migration and the Office for the Coordination of Humanitarian Affairs. In order to combine the datasets, I previously standardized in Excel the questions across the five surveys. Additionally, I deleted (majorily) empty columns, columns that don't exist across all five datasets, and columns with identical/highly similar information. I also divided columns with more than one data point per cell into the respective amount of individual cells.
The original datasource: Humanitarian Data Exchange
The data analysis involved extensive cleaning of the dataset derived from the five surveys. The final dataset contains 861 entries (=locations with differing number of households) and 48 features. The machine learning algorithm applied is KMeans for clustering and involves Principal Component Analysis (PCA).
During PCA, 10 features were identified to weight most allowing to identify specific drivers for mobility and displacement in the region.
Visualization of the main findings: Tableau Story
**SUMMARY OF FINDINGS:**
1. In one third of locations (village unit or urban settlement/town) in the region surveyed, only 25% of inhabitants have access to cooking water, the same applies to drinking water.
2. In on average 40% of lo …