**10. International development programs that use PAAs**

From a geopolitical perspective, I have also included case studies on themes that are universally applicable with specific emphasis on select themes that significantly contribute in making the world a better place and hence promoting a better quality of life.

### **10.1 Natural disaster programs**

The concept of predictive analytic algorithms has been implemented in the analysis of big data regarding past natural disasters and used to predict future incidences [23]. An example of an incident that provided data for fighting natural disasters is the earthquake that occurred in Haiti in 2010. Crowdsourcing has been used to obtain real-time images of disasters such as earthquakes while big data approaches in artificial intelligence (AI) have been used to determine meanings in messages such as SMS that were generated during the occurrence of natural disasters.

The processes involved the installation of machine learning language followed by the creation of an algorithm that enables the performance of mathematical analyses such as regression analysis and providing the output that can be interpreted to estimate the likelihood of occurrence of a similar incident such as another earthquake in the future [33]. The analytical procedures performed involve the input of information pertaining to disasters such as the magnitude of an earthquake, time of occurrence, and region into the machine language. The machine language performs an analysis of mathematical processes such as linear regression and multiple regressions to provide statistical coefficients that can be used to predict future disasters.

### **10.2 Poverty eradication program**

Predictive analytics have been used by the World Bank (WB) in poverty eradication initiatives such as the collection of information of affected areas, the analysis of the number of people who need relief services, and the relationship between their status with infectious diseases. This is in accordance with the WB objective of eradicating poverty by the year 2050. Countries conduct household surveys and provide WB with information used to classify the population according to the level of poverty [25].

The processes involve the creation of a machine language that enables input of raw data such as the economic statuses of families. Data from statistical offices in various countries are input into the machine learning language that has been designed in a customized fashion to enable the stratification of families according to their gender, age, income levels, geographical location, race, or culture. The program has commands that enable the quick calculation of statistical measures such as linear regression or multiple regressions to provide coefficients that enable the prediction of poverty levels in the future [2]. The machine learning language has also been designed in a manner that enables the transfer of data from mobile phones to the program for analysis. This program has been implemented to measure the economic status of people in Togo, Tanzania, and Tajikistan to provide outputs that enable prediction of poverty status in the future. A similar program has been used by the WB in the measurement of the movements of nomadic people in Somalia to predict future migration patterns.
