Webinar #7 - Mining Software Development History with Deep Learning

GDG Cloud Singapore
Thu, Jul 16, 2020, 7:30 PM (GMT+8)

20 RSVP'ed

About this event

Webinar will consists of 3 main sections:

1. Human RSS

A short sharing session of latest news of products/services coming out from Google Cloud

2. Main topic - Untold Story of Your Code: Mining Software Development History with Deep Learning

Speaker: Mikhail Filippov

Mikhail Filippov is chief scientist at Quod AI and a research fellow at University College London. At Quod AI he is applying natural language processing, deep learning and statistical models to explain source code in plain English.

Mikhail holds a PhD in physics from Nanyang Technological University. As a mathematical physicist he studied the dynamics of complex systems and has built mathematical and AI models for tsunami prediction, tropical atmosphere, the housing market, historical information and visual cortex of the human brain. His models have been used by NASA & the Monetary Authority of Singapore.


Any piece of software, a final product that we use is just a tip of an iceberg, which is a complex process of software development. We can mine this iceberg for the insights into how the project evolved into its final stage as well as how the developers team did work and collaborate.

To do that we have to equip ourselves with classic and advanced machine learning techniques: from clustering and segmentation to deep learning and NLP.

In this talk we will take a look at what insights could be extracted from the software development history, how similarities between natural and programming languages can be useful and what Google Cloud Platform tools can help us on this treasure hunt.

Additional References for the talk:

Some key takeaway from the session:
- How machine learning could be used on code?
- Which google tools (apart from Tensorflow) could be useful for that and why?
- How can machine learning on code help with productivity of a software developers team?

Please proceed on to the live links on youtube/facebook to watch the live stream, or watch it later at your own pace
Youtube link: https://www.youtube.com/watch?v=PUKTCLwTqN0