Supporting Artificial Intelligence with geospatial data in Nigeria: GRID3 joins Data Science Nigeria’s AI Bootcamp
From 18 to 23 November 2019, GRID3 Nigeria joined Data Science Nigeria’s Artificial Intelligence (AI) Bootcamp in Lagos to train the next generation of AI experts on geospatial analytics.
Raising one million Artificial Intelligence talents across Nigeria within the next 10 years: this is the vision of Data Science Nigeria. To provide aspiring and established data scientists with requisite world-class skills and industry-relevant knowledge to support transformational solution development and employability in Nigeria, the non-profit organisation hosts the AI Bootcamp in Nigeria each year, uniting close to 200 delegates from across the country.
GRID3 Nigeria contributed its expertise by introducing delegates to the power of geospatial analytics for sustainable development. Before the official kick off, the GRID3 Nigeria team delivered presentations at the AI Summit and the AI Executive Masterclass pre-Bootcamp events. Focusing on the link between geospatial data and AI applications, the presentations emphasised the value of gridded population data for development and businesses, with particular attention given to applications related to health, education and financial inclusion.
GRID3 Nigeria opened the AI Bootcamp with a session on the “Introduction to Geospatial Analytics”. The team presented the link between Geographic Information Systems (GIS) and AI, data analysis workflow, GIS data models, and the range of applications offered by this method. Emmanuel Jolaiya (@jeafreezy), one of the AI delegates, commented on Twitter:
I'm super excited to hear that @grid3global is at the @DataScienceNIG bootcamp.
Now that participants(Data enthusiasts) have been introduced to #GIS.I smell amazing & disruptive geospatial Innovations after the bootcamp. :) https://t.co/nDkOwLPilw— Emmanuel Jolaiya (@jeafreezy) November 20, 2019
Later in the week, the team ran a parallel, in-depth session on the “Production and Use of Gridded Population Data”, introducing delegates to the GRID3 bottom-up modelling approach and the gridded population datasets. Participants took the opportunity to interact with GRID3 data through an interactive data demo activity:
Yayyy!!! My first R code @grid3global hands on @DataScienceNIG #AIBootcamp pic.twitter.com/3q0QusxYyH
— Lawal Adeola (@azlawal_lawal) November 22, 2019
The team from @grid3global has really nailed it. #DSN_AI_Bootcamp2019#1million_AI_talents_in_10_years pic.twitter.com/60s5qEXZMN
— Faruq Ahmed (@faruqjada) November 22, 2019
Using RStudio Cloud and GRID3 data on “Population data for women age 14-49”, “Ward and state boundaries” and “Health facility locations”, delegates were asked to assess coverage of health facilities for maternal health in Kaduna State.
Ahmed Olanrewaju commented on Twitter:
What a life time opportunity to jump right into GeoSpatial Data Analysis. All thanks to @DataScienceNIG #AIBootcamp for making this to happen. @grid3global thanks . @IbadanRusers @gbganalyst , there is work for us . I have installed packages like rjags, runjags on my @rstudio pic.twitter.com/jbWHjTGPmt
— Ahmed Olanrewaju (@abono2000) November 22, 2019
At the end of the workshop, delegates were given the opportunity to participate in a data challenge competition to find innovative ideas around the use of gridded population data in Nigeria – Watch this space for the results.
GRID3 is committed to mapping a path to sustainable development for everyone. The AI Bootcamp proved to be a unique opportunity to meet many data enthusiasts eager to learn new skills and integrate new data and methods into their work.
Only a few weeks later, former Bootcamp participant Victor Irekponor and his team, won a data hackathon, integrating new GIS skills from GRID3 training. Read his testimonial below:
Friday, 29th Nov – Sunday, 1st Dec 2019 was spent with my team at the data hackathon organised by the African FIntech Foundry Powered by Access Bank Plc.
My team was tasked with developing a machine learning model that significantly reduces Non-Performing Loans for individuals and also an alternate credit scoring model based on the customer’s historical and transactional behaviour.
Unlike other teams, we shifted our main focus from just building a machine learning model to providing implementable and actionable business insights. Geospatial analysis added significant value to our presentation, and was likely the reason that we won. Many thanks to GRID3, as they introduced us to the concepts behind geospatial analysis at the DSN Bootcamp held 18th – 24th Nov 2019.
We reverse geocoded the addresses of the Access Bank customers given to us in the dataset, to get the points in latitude and longitude, which we then plotted on the map of Nigeria to show Access Bank executives where exactly their defaulting customers are clustered at both national and regional levels, as well as their distribution.
This was one of our key selling points. In the end, my team emerged First Place Winner at the Hackathon. We were the only team that utilised the strengths of geospatial analysis.