The 1st Big Data Summit and 2nd Data Science Bootcamp were held between 12 and 15th October at the Oriental Hotel, Lekki and the Peninsula Resort, respectively. The two events were well attended by the leading experts, players, and enthusiasts in the emerging big data ecosystem, from the industry, developer and academic communities. Below are some reflections of the two events, hosted by Data Science Nigeria, a vision of the MTN’s executive, Mr Olubayo Adekanmbi.
Keynote by Dr Yemi Kale, Statistician General of the Federation
- A rapidly growing universal truth in international business today is that all roads lead to data.
- The sexiest and most sought-after job in the 21st century is data science.
- Policy-makers have ignored official data and made up their own, irrespective of the fact that it was largely "guesstimating".
- We're now in an information-driven economy where data is the main currency.
Welcome Speech by Bayo Adekanmbi, Convener, Data Science Nigeria
- Our focus at Data Science Nigeria has been on capacity building and world-class mentorship, and our vision is to see Nigeria become the hub for data experts and machine learning in the next 10 years.
- We must change our data analytics practice from descriptive and predictive to prescriptive and cognitive.
- Data Science Nigeria already has three major big data products on local language, the emoticon analysis of tweets, and an integrated marketing dashboard that uses daily weather/inflation/national sentiment to predict sales.
Industry perspective by Olumide Olayinka, Partner and Head of Analytics Practice, KPMG
- Machine learning and Artificial Intelligence are not taking away our jobs. It's the rise of humans to create better jobs.
- In God we trust, but every other man must bring data.
- The faster the information is provided, the faster decisions can be made. Data allows us to see patterns, understand changes, and make accurate predictions.
Executive Perspective by Uzoma Dozie, CEO, Diamond Bank
- The organization that makes the right decisions all the time wins.
- Twenty years ago, accountants were important as they look at historical records. Today, data scientists are important because we want algorithms that understand the future and can predict how to meet the needs of customers.
- If your data doesn't address your organisation’s growth concerns, cost concerns, and risk concerns, it's useless.
Panel Discussion 1: Roadmap for Nigerian Big Data Development
We have the local capacity for massive data analytics in this country. We must not look down on ourselves. - James Agada, CEO, CWG
Nigerians don't want to know all the background analytics. They want to see things work! A good use of big data makes the developer look like a magician. - Chika Nwobi, CEO, L5Lab
We will not have a tech ecosystem until we focus on developing fundamental STEM pillars. We cannot be excited about being software consumers without leveraging core research through our universities to drive breakthrough innovations. - Emeka Okoye, CEO, Cymantiks
Asking the right questions helps us to capture the insights from data. Before we consider the applications, let's look into the fundamentals. - Yomi Akinyemi, Senior Manager, Analytics, KPMG
Big data can be simplified to data on demand via a cloud-based approach in which you get the best of what you want at the lowest possible fees that you can afford. - Wale Olokodana, Director, Enterprise Commercial Segment - Microsoft
Data is breaking away from the IT department and becoming an integral part of every department in a company. We must build a data-driven economy where data solves the problems of transactions without any limitations. - Babafemi Ogungbamila, CIO, Interswitch
Unstructured data from local Nigerian sources can be properly analysed with many currently existing solutions using natural language algorithms, and the necessary customisation can be made to make technology serve our local nuances with ease. - Chidiebele Ifediora, Technical Sales Leader, IBM
Panel Discussion 2: Practical local case studies on big data solutions
To build the Nigeria big data ecosystem, we need passionate learners, world-class mentors, learning communities, and local solutions. - Nadayar Enegesi, Andela
A smart city will only be possible by using big data. When we are able to make sense of the multiple data around us, we will be able to track diseases, and make the nation safer, better, and well-structured for our collective well-being. - Bernard Ewah, CCO Keoun.ng
Artificial intelligence and machine learning will disrupt advertising and media, but this does not mean it will take our jobs. Steve Baba-Eko, CEO X3M
There is a whole concept of individualised learning pathway that can be built to support a personalised education per individual based on an intelligent use of big data to understand variation in our learning style, approach and cognitive capabilities. There is no one size fits all. - Dr Tunji Adegbesan, Founder, Gidimo
With advanced analytics, we can score drivers, predict their risk level, and even use our data to advise insurance companies. - Mr Foh Patrick, Software Enginner, Maxdotng
At Paylater, we take your data, make a prediction, and credit your account within minutes if you're eligible. - Ngozi Dozie, Director, OneFi
Nigerians don't like maps; they want to ask someone. We are an agent-based economy, hence the reason why we created Lara.ng as local alternatives to Google maps. - Samuel Odeloye. Co-Founder Lara.ng
We must go back to the primary schools and fortify our science, technology, engineering and math (STEM) education. The kids have capacity to learn faster by being exposed to data science earlier. - Dele Tejuoso, Founder, Wifi Combat
Case study presentation of Zenvus, precision agriculture solution by Professor Ndubuisi Ekekwe
- Zenvus’ precision farming intelligence solution is an intelligent solution for farms that use proprietary electronics sensors to collect soil data, such moisture, nutrients, pH, and send them to a cloud server via GSM, satellite or Wifi. Algorithms in the server analyse the data and advise farmers on what to do to maximize yields.
- In product development, go beyond the needs and expectations of your customers— go to the level of your customers’ perceptions.
- Let us promote the need to learn the fundamentals so that we do not become limited by existing platforms. Let us learn the mathematics of these things, and not become a victim of the final products’ consumers.
- You have to believe that you are good. If you don't believe you are good, then you're not.
Industry Showcase of Microsoft Azure by Damola Solanke, Apps & Infrastructure Lead at Microsoft
- There is a lot of intelligence in the Microsoft platform to help you get better insights from your data.
- Data is good, but data with meaning helps a business to grow faster and smarter.
- Microsoft Azure makes it easy for companies to access world-class tools without the traditional cost of software and hardware.
Four-day Data Science Bootcamp Summary
Four days of world-class learning, seven PhD level data scientists as on-site tutors and dial-in instructors, three on-going Hackathons, and 1 female winner (Adeola Dorcas, a level 300 computer science undergraduate of the Obafemi Awolowo University, Ile-Ife.
Dr Raphael Yemitan of PwC explored the fundamental principles of feature engineering in a data science project. Dr Sulaimon Afolabi, a leading data scientist from South Africa, led a step-by-step class into the fundamental theories of data science from statistical and machine learning perspectives.
Corne Nagel, Chief Data Scientist with OneFi, explored both the bottom-up and top-down approaches of model development using an exciting showcase of Zeppelin, and the hands-on use of automatic ML with DataRobot. Zeppelin is a solution which allows users to combine R, Scala, and Python in a single, seamless pipeline on the fly. He also kicked off the second hackathon on credit risk scoring.
Wale Akinfaderin, a doctoral researcher and expert data scientist dialled in from the USA to discuss feature engineering, data preparation, overfitting, and managing missing data (including mean, median, mode, regressed value or even nearest neighbour values). Professor Raj Krishnan of the Illinois Institute of Technology in Chicago and a Microsoft leading expert in Microsoft Azure led a hands-on immersion in machine learning using the cloud-based Azure platform.
Dr Johnson Iyilade from Canada explored how we have moved from a system-centric to a user-centric connected data architecture, with practical explorations of the major big data technologies (Hadoop, HDFS, MapReduce, etc.). Dr Ponmile Oloyede, a computational physicist and USA-based finance modelling expert showed many practical use cases of big data in everyday life and introduced Rattle as a simple R interface.
Ladi Aduni, an Associate Director with KMPG led the briefing on the KPMG Segmentation hackathon at the bootcamp. The hackathon required participants to use unsupervised learning to cluster banking customers. The best participants in this hackathon will qualify for internships and possible job placements at KPMG. The project was extended beyond the bootcamp for maximum participation.
Ngozi Dozie of OneFi was at the bootcamp discussing how to build industry-ready algorithms that can solve real-world problems. For example, he spoke on how weather forecasting correlates with loan default. When the weather is bad, people default more on their loans.
Usoro Usoro, General Manager of Digital Financial Services engaged the bootcamp participants on how to leverage the robust intelligence of big data algorithms to unravel local opportunities in payments and lending within the emerging Nigerian financial tech space.
Mrs Oby Ezekwesili, former Minister o Education followed the event real-time on Twitter and challenged the team to drive the structural shift that the 2nd Machine Age has made imperative with a strong focus on raising girls’ STEM education.
The event had a gala night, during which the top performers on the Kaggle competitions, Mr and Ms Algorithms (individuals who epitomise values of leadership, excellence and team-spirit), PhD participants, bootcamp instructors, and volunteers were recognized with gifts and accolades.
Bootcamp participants came from far and near; over 65% were from outside Lagos. Major clusters of attendees include Owerri (FUTO), Kaduna (Ahmadu Bello University, ABU Zaria), the southwest (Ibadan, Ile Ife, Ogbomosho, Abeokuta and Ilorin) among other locations.
The two events were supported by Microsoft, 4Afrika, Interswitch, Diamond Bank, OneFi, L5Lab, Keoun.ng, KPMG, ProShare Nigeria, Nigerian Bureau of Statistics, DataCamp, BusinessDay, CFATech.ng, BrandCrunch and the Machine Intelligence Institute of Africa.