
Often underestimated in AI: The quality of data and modeling
A Talk by Martin Treder (Business Data Consultancy, ReturnOnData.biz)
About this Talk
Next to people, Data Science depends on algorithms, tools, data, modelling and hardware.
Algorithms, tools and hardware (GPUs, cloud solutions) all have progressed significantly during the past years. Data Scientists know where to go. Libraries like TensorFlow and PyTorch have made life easy for them. But how good is our data quality, and how scientific is our modeling of real-life problems?