Please log in to watch this conference skillscast.

At Onfido, they apply computer vision and machine learning techniques to the detection of physical and digital fraud in identity documents.

When it comes to fraud detection, they’re often dealing with a problem for which they have hugely unbalanced datasets. Moreover, they only understand what they've seen before; if they notice something different, how do they know if it’s really fraudulent, and how confident can they be with that decision?

Join Jacques for this lightning talk as he explores the problem of fraud detection, highlight some of the biggest challenges that they face in this area and cover some of Onfido's efforts to circumvent these challenges, touching on topics of anomaly detection and data generation.

YOU MAY ALSO LIKE:

• Fast Track to Machine Learning with Louis Dorard (in London on 15th - 17th July 2019)
• Real-time Systems with Spark Streaming and Kafka (in London on 23rd - 24th September 2019)
• Infiniteconf 2019 - A one-day community celebration of Big Data, Machine Learning and AI (in London on 4th July 2019)
• CloudNative London 2019 (in London on 25th - 27th September 2019)
• Causal inference and the data-fusion problem (in London on 18th June 2019)
• The wonders of IBM: Watson, artificial intelligence, Swift coding and databases! (in London on 19th June 2019)
• AI Auditing Framework (SkillsCast recorded in June 2019)
• Applying Monte Carlo Tree Search (MCTS) to the Protein Folding (SkillsCast recorded in June 2019)