Microsoft, KPMG and UNESCO jointly organized the first two virtual hackathons on artificial intelligence called Imagine Cup Junior 2022. The aim of the events is to dispel myths about AI and to use it to solve real-world problems. Team Kazakhstan took 1st and 2nd place in different sections of the event series.
Twenty-one teams from nine countries participated in the first hackathon of the series, which focused on using AI to preserve the Earth. Each team consisted of 5 students and 1 faculty member. Participants used AI to solve major environmental and sustainability issues facing our planet. In order for the idea to be applicable in the real world, students had to predict problems, use creative thinking, and show particular attention to detail.
The Unicode team from Kazakhstan took 2nd place. The team identified that some road construction projects, such as China's Belt and Road project, were a major threat to tiger populations, and developed a plan to work with development companies to identify alternative routes that would protect important animal habitats.
The team's solution uses an image classification model to record sightings and locate tigers by analyzing satellite imagery and then selecting optimal alternative routes.
First place in the hackathon went to the Data team from Latvia with a solution to conserve the world's bat population, while the Bulgarian teams Quadcore and Yottabyte shared third place with conservation solutions.
The second event, the Imagine Cup Junior Virtual AI Hackathon, a section for girls, brought together 22 teams from different schools in Kazakhstan, Pakistan, Saudi Arabia, Bulgaria, the UK and the UAE. The event aimed to address the significant gender differences in Science, Technology, Engineering and Mathematics (STEM) fields by supporting young women who have expressed an interest in AI.
Kazakhstan's Cognitive team took 1st place with a project aimed at eliminating bias and human error in domestic violence claims. Their solution uses a two-step process to determine the severity of an emergency call by analyzing the Police Department's criminal database as well as service records on domestic violence.
The AI processes thousands of data points to classify calls based on the likelihood of domestic violence and prioritizes emergencies with high accuracy. The girls' hackathon section helped participants develop specific skills for working with statistical data, computer vision (a field of AI that allows computers and systems to extract important information from digital visual data), and natural language processing skills.
They also learned about data collection and cleaning, model selection and statistical testing, and worked with regression, decision trees (decision tools used in machine learning, data analysis and statistics) and random forests (machine learning algorithms that use sets of decision trees).