Abstract: Machine learning applications are new in the software development landscape, and tend to be hard to build. As Google noted in an article (source : https://papers.nips.cc/paper/5656-hidden-technical-debt-in-machine-learning-systems.pdf), it is mainly because the application is much broader than the model itself. Surprisingly though, Machine Learning applications follow a double Pareto’s law. On the one hand, 80% of the time spent on building those applications deals with machine learning problems whereas 20% of the remaining time is spent on integrating the model to a running application. On the other hand, only 20% of the code lines are specific to machine learning ; the vast rest is about integration and run.
I would like to first explore the foundations of this trend, to then show why it kills machine learning application development and sustainability.
In order to illustrate the tips and tricks of shipping a deep learning model to production, I would use a live demo of a model designed to recognised car drawings via the camera of a Raspberry Pi. It would allow me to. Furthermore, by identifying the main stages of the application life cycle (training, deploying and using), I will lay the emphasis on the common mistakes one does not bear in mind to make a successful machine learning product.
Bio: Constant is strongly interested in the creation of value out of data and helps those who believe in such a potential by accelerating their transition toward a data driven company. In order to address these new problematics, he focuses on mastering every skill of a complete Data Geek : architecture expertise (data, applications, network), data science mastering (statistical learning, data visualisation, algorithmic theory), customer and business understanding (model prediction consumption, business metrics, customer needs).
Constant has been working for about two years for OCTO Technology. He is an an expert in the industry sector and works on several types of mission, ranging from predictive maintenance of production site, to prediction of critical KPIs in video games, via real time monitoring of manufacturing devices. Prior to joining OCTO, Constant was working as a researcher in data61 (formerly known as NICTA), the best research institute in ICT in Australia on applying Machine Learning to profile GUI users and provide the best amount of information to help them make a decision based on a machine learning prediction.