top of page
  • Noé Carème Fouotsa Manfouo

Data revolution

Wondering what the Data Science hype is all about? It is a revolution that is changing the world with tangible impact. Organisations consider their data science teams as one of their biggest assets in today’s competitive environment. Data science finds applications in finance, retail, healthcare, manufacturing, sport, communication etc.

Glassdoor, one of the major hiring interfaces in Europe has ranked data scientist jobs as the best job for 2017, both in terms of salaries and market demand. The United States alone will need 190 000 data scientists by 2018, with an average salary of $120 000. Besides the buzz around data science as a career, you might actually wonder what exactly it is. According to Vansant Dhar, it is an interdisciplinary field about processes and systems to extract knowledge or insights from data in various forms, either structured or unstructured, which is a continuation of some of the data analysis fields such as statistics, data mining, and predictive analytics, similar to Knowledge Discovery in Databases (KDD).

South Africa has set a target of reaching a critical mass of data scientists by 2025 in order to efficiently answer the decision support challenges raised by the globalisation of economies. The strategy provides scholarship funds to attract Masters and PhD candidates. The five days workshop organised at the African Institute of Mathematical Sciences (AIMS) was to raise such awareness, and one of our Surgor members, Noé Manfouo, was happy to attend. The workshop was organised in four main streams, namely (1) theoretical foundations of data sciences; (2) computer lab sessions; (3) real world applications and finally (4) special tutorial sessions.

Theoretical foundations of data sciences entails more than intermediate level in linear algebra, matrix calculus, heuristic optimization, statistics, probabilities and functional analysis. During the computer lab sessions, theoretical foundations was used to implement the main machine learning algorithms in Python, namely the Artificial Neural Network (ANN), the Recurrent Neural Network (RNN), the Long-Short Term neural Network (LSTM), the Convolutional Neural Network (CNN), based on certain industry issues. Some case studies included stock market prices forecast, speech and image recognition.

The real world application module of the workshop included applications to supply chain management, security, healthcare, wildlife protection and cosmology. Finally, the special tutorial sessions were used to provide more opportunities to participants to improve their skills, in a case-to-case basis.

A broader perspective on machine learning and artificial intelligence techniques show that the field has grown so fast to reach enormous potential, which brings a mixture of hope and worries for the future, just as nuclear physics in the 30s. As a matter of fact, machine-learning algorithms was able to beat a human being for the first time on a free style chess challenge in 2017. Russia was able to build a so-called invincible missile able to overcome the latest anti-missile protections technology. This can only contribute to the feeling of insecurity. The question of ethics in data science, machine learning and artificial intelligence must be seriously considered.

To conclude, the workshop was a very interesting training session and we look forward in both practicing the knowledge acquired and attending the next session organised in August 2018.

13 views0 comments

Recent Posts

See All
bottom of page