Combining formal methods and machine learning

Duration: 1 hour 9 mins

Share this media item:

Embed this media item:

<iframe width="_width_" height="_height_" src="https://sms.cam.ac.uk/media/2994029/embed" frameborder="0" scrolling="no" allowfullscreen></iframe>

Choose size:

About this item

Available Formats

About this item

Combining formal methods and machine learning's image

Description:	Formal methods offer rigorous techniques for reasoning about systems, while, more recently, data-driven approaches based on machine learning have become popular. The two techniques are complementary, and this talk explores ways in which thay have been combined together.


Created:	2019-05-30 16:07
Collection:	Annual Wheeler lectures
Publisher:	University of Cambridge
Copyright:	Prof. Jane Hillston
Language:	eng (English)
Keywords:	Formal methods; Machine learning; Systems analysis;


Abstract:	Quantitative formal methods, such as stochastic process algebras, have been successfully applied in a number of application domains over the last 20 years. They offer rigorous techniques for asking questions about the dynamic behaviour of systems. In the last decade more data-driven approaches to system analysis, based on machine learning have gained prominence. Yet the two approaches have complementary strengths and weaknesses and should not necessarily be thought of as competing. In this talk I will talk about two pieces of work in which we have sought to combine machine learning techniques into a formal modelling framework.

Abstract:

Quantitative formal methods, such as stochastic process algebras, have been successfully applied in a number of application domains over the last 20 years. They offer rigorous techniques for asking questions about the dynamic behaviour of systems. In the last decade more data-driven approaches to system analysis, based on machine learning have gained prominence. Yet the two approaches have complementary strengths and weaknesses and should not necessarily be thought of as competing. In this talk I will talk about two pieces of work in which we have sought to combine machine learning techniques into a formal modelling framework.

Available Formats

Format	Quality	Bitrate	Size
MPEG-4 Video	640x360	956.99 kbits/sec	483.64 MB	View
WebM	640x360	309.56 kbits/sec	156.45 MB	View
iPod Video	480x270	521.41 kbits/sec	263.51 MB	View
MP3	44100 Hz	250.7 kbits/sec	126.70 MB	Listen
Auto *	(Allows browser to choose a format it supports)