The course aims to illustrate the main techniques of machine learning, the structure of the process and its use in humanitarian projects.
Gnucoop's many years of experience in the development of mobile web platforms for the monitoring of cooperation interventions has allowed us to acquire an excellent background in the management of the collection and construction of data analysis processes related to interventions. AI&ML system can be of great help in many development contexts. For example, through a face recognition algorithm we created a mobile app that can be used by teachers to register school attendance in Burkina Faso. Moreover, in our data collection system DEWCO, data validation is performed not only through standard logic but also by learning from the natural correlation within existing data. In this section we will explore the basics of Python for Data Analysis. We will use the popular Jupyter Notebook interface for Python for learning the most commonly used machine learning techniques to tackle the real world data analysis challenges. It is a common knowledge that cleaning and normalizing the data requires the most amount of time for an analysis project even to start. We will see how machine learning can help us in cleaning and categorizing our dataset, see some Clustering Algorithms for an exploratory analysis of our data, and Decision Tree Classifiers for understanding how the most relevant aspects of our data could be found.
- Machine Learning for Data Cleansing;
- PICAPS - using data analysis and data mining methods to describe the socio-economic situation of families in some villages in Burkina Faso, with the intention of understanding in which environments children are more likely not to attend school.
The course will last 4 weeks and it will be held in English. The online component includes tutoring activities during the course and access to an online training platform for sharing materials.
For more information firstname.lastname@example.org