
REGISTRATION & LIGHT BREAKFAST

WELCOME
Fabrizio Silvestri - Facebook
Embeddings in the Real World: Two Case Studies
We present two novel embedding mechanisms that are derived for two particular applications in search: spellchecker and queries. The two novel embeddings are designed to improve the quality of the underlying services with the constraint of not increasing the computational time necessary to process queries. We show how we tackled this problem along with some preliminary results on test datasets.
Fabrizio Silvestri is a Software Engineer at Facebook London in the Search Systems team. His interests are in web search in general and in particular his specialization is building systems to better interpret queries from search users. Prior to Facebook, Fabrizio was a principal scientist at Yahoo where he has worked on sponsored search and native ads within the Gemini project. Fabrizio holds a Ph.D. in Computer Science from the University of Pisa, Italy where he studied problems related to Web Information Retrieval with particular focus on Efficiency related problems like Caching, Collection Partitioning, and Distributed IR in general.


THEORY & LANDSCAPE


Andreas Damianou - Machine Learning Scientist - Amazon
Probability & Uncertainty in Deep Learning
Andreas Damianou - Amazon
Probability & Uncertainty in Deep Learning
In this talk I will motivate the need for introducing probabilistic and Bayesian flavor to "traditional" deep learning approaches. For example, Bayesian treatment of neural network parameters is an elegant way of avoiding overfitting and "heuristics" in optimization, while providing a solid mathematical grounding. I will also highlight the deep Gaussian process family of approaches, which can be seen as non-parametric Bayesian neural networks. The Bayesian treatment of neural networks comes with mathematical intractabilities, therefore I will outline some of the approximate inference methods used to tackle these challenges.
I completed my PhD studies under Neil Lawrence in Sheffield, and subsequently pursued a post-doc in the intersection of machine learning and bio-inspired robotics. I've now moved to Amazon as a machine learning scientist, based in Cambridge, UK. My area of interest is machine learning, and more specifically: Bayesian non-parametrics (focusing on both data efficiency and scalability), representation and transfer learning, uncertainty quantification. In a recent strand of work I seek to bridge the gap between representation learning and decision-making, with applications in robotics and data science pipelines




Shubho Sengupta - AI Research - Facebook AI Research (FAIR)
Sequence to Sequence Networks for Speech & Language Translation
Shubho Sengupta - Facebook AI Research (FAIR)
Systems Challenges for Deep Learning
Training neural network models and deploying them in production poses a unique set of computing challenges. The ability to train large models fast, allows researchers to explore the model landscape quickly and push the boundaries of what is possible with Deep Learning. However a single training run often consumes several exaflops of compute and can take a month or more to finish. Similarly, some problem areas like speech synthesis and recognition have real time requirements which places a limit on how much time it can take to evaluate a model in production. In this presentation, I will talk about three systems challenges that need to be addressed so that we can continue to train and deploy rich neural network models.
Shubho is now working on AI Research at FAIR. He was previous a senior research scientist at Silicon Valley AI Lab (SVAIL) at Baidu Research.
I am an architect of the High Performance Computing inspired training platform that is used to train some of the largest recurrent neural network models in the world at SVAIL. I also spend a large part of my time exploring models for doing both speech recognition and speech synthesis and what it would take to train these model at scale and deploy them to hundreds of millions of our users. I am the primary author of the WarpCTC project that is used commonly for speech recognition. Before coming to the industry, I got my PhD in Computer Science from UCDavis focusing on parallel algorithms for GPU computing and subsequently went to Stanford for a Masters in Financial Math.



Jonas Lööf - Deep Learning Solution Architect - NVIDIA
Considerations for Multi GPU Deep Learning Model Training
Jonas Lööf - NVIDIA
Considerations for Multi GPU Deep Learning Model Training
As the amount of available data grows, deep learning products and applications profit from the use of bigger and bigger models. To effectively train these models in a timely manner it is necessary to parallelize training over multiple GPUs and multiple machines. In this talk, we explore what is needed to efficiently achieve this kind of scaling, both from a hardware and library perspective, but also from the perspective of the end user. Methods of multi GPU parallelization are described and discussed, and best practices are presented. Furthermore, the choice of deep learning framework, and its impact on multi GPU training is discussed, along with the resulting tradeoff between flexibility and engineering effort. Finally, examples and benchmarking results are described and discussed, showing the possibility of near-linear scaling of training time both to multiple GPUs and multiple machines.
Jonas Lööf is a Deep Learning Solution Architect at NVIDIA, where he helps guide customer decision making on both hardware and software in their deep learning projects. Before joining NVIDIA, Jonas has worked in research and development, applying deep learning in the fields of speech recognition and natural language processing, both in a startup environment and the corporate world. Jonas holds a doctoral degree in computer science from RWTH Aachen University, where he worked on acoustic model adaptation for speech recognition.



COFFEE
REINFORCEMENT LEARNING & UNSUPERVISED LEARNING
Marta Garnelo - DeepMind
Representations for Deep Learning
Current deep learning algorithms have achieved impressive results on a variety of tasks ranging from super-human image recognition to beating the world champion at the game of Go. Despite these successes deep learning algorithms still suffer from a variety of drawbacks: they require very large amounts of training data, they lack the ability to reason on an abstract level and their operation is largely opaque to humans. One way to overcome these problems is by creating models that form useful representations which exhibit beneficial properties such as disentanglement, the ability to generalise etc. This talk focusses on recent work that has addressed this issue for deep learning models.
Marta is a research scientist at DeepMind and also currently halfway through her PhD at Imperial College London under the supervision of Prof Murray Shanahan. Her research interests include deep generative models and reinforcement learning, in particular finding meaningful representations using the former to improve the latter.


SEARCH & REASONING
Fabrizio Silvestri - Facebook
Embeddings in the Real World: Two Case Studies
We present two novel embedding mechanisms that are derived for two particular applications in search: spellchecker and queries. The two novel embeddings are designed to improve the quality of the underlying services with the constraint of not increasing the computational time necessary to process queries. We show how we tackled this problem along with some preliminary results on test datasets.
Fabrizio Silvestri is a Software Engineer at Facebook London in the Search Systems team. His interests are in web search in general and in particular his specialization is building systems to better interpret queries from search users. Prior to Facebook, Fabrizio was a principal scientist at Yahoo where he has worked on sponsored search and native ads within the Gemini project. Fabrizio holds a Ph.D. in Computer Science from the University of Pisa, Italy where he studied problems related to Web Information Retrieval with particular focus on Efficiency related problems like Caching, Collection Partitioning, and Distributed IR in general.




Sebastian Riedel - Reader - UCL / Bloomsbury.ai
Towards Teaching Machines to Read and Reason
Sebastian Riedel - UCL / Bloomsbury.ai
Towards Teaching Machines to Read and Reason
We are getting better at teaching machines how to answer questions about content in natural language text. However, progress has been mostly restricted to extracting answers that are directly stated in text. In this talk, I will discuss our work towards teaching machines not only to read, but also to reason with what was read. We investigate two complementary approaches. In the first, we incorporate reasoning into neural models by enabling them to manipulate memory in a hierarchical sequence of steps—just like a virtual machine that executes a program. In the second, we take black box models (neural or not) and keep their structure as is. Instead, we develop rewards and penalties that encourage the model to behave as if it reasons. I will show the effectiveness of these approaches on various benchmark datasets, including relation reasoning and solving math word problems.
Sebastian is a Reader at University College London, leads the UCL Machine Reading group and is Co-Founder and Head of Research at Bloomsbury AI. He is an Allen Distinguished Investigator and received an $1M award from the Paul Allen Foundation to 'move the needle' towards answering broad scientific questions in AI. Sebastian is generally interested in the intersection of Natural Language Processing and Machine Learning, and particularly interested in teaching machines to read and to reason with what was read.



LUNCH
NATURAL LANGUAGE PROCESSING


Nikolaos Aletras - Applied Scientist - Amazon
Labelling Topics Using Neural Networks
Nikolaos Aletras - Amazon
Labelling Topics Using Neural Networks
I am an Applied Scientist at Amazon, Cambridge, UK. Prior to that, I worked as a Research Associate at the Department of Computer Science at UCL and I completed a PhD in NLP at the Department of Computer Science of the University of Sheffield. My main research interests are in Natural Language Processing and Machine Learning. More specifically, I'm interested in applying statistical methods for detecting and interpreting the underlying topics in large volumes of text data. I also develop methods to analyse text and uncover patterns in data to solve problems in other scientific areas such as social and legal science.
Much of the information in large digital libraries is often stored in an unstructured way and is not organised using any automated system. That is usually overwhelming for users in a way that makes it difficult to find specific information or explore such data collections. A particular set of unsupervised statistical methods, namely topic models have been extensively used in Natural Language Processing and Information Retrieval for automatically analysing and organising document collections. Topics generated by topic models are typically presented as a list of terms. Automatic topic labelling is the task of generating a succinct label that summarises the theme or subject of a topic, with the intention of reducing the cognitive load of end-users when interpreting these topics. In this talk, I will present neural network approaches to labelling topics with text and images showcasing their effectiveness on providing meaningful representations of the topics.




Elena Kochkina - PhD Student - University of Warwick
Opinion Mining Using Heterogeneous Online Data
Elena Kochkina - University of Warwick
Opinion Mining Using Heterogeneous Online Data
Understanding public opinion is important in many applications, such as improving company's product or service, marketing research, recommendation systems, decision and policy making and even predicting results of elections. Social media is a very powerful tool to transfer information and express emotions for users and a rich source of data that enables researchers to mine public opinion. In this talk I will describe different lines of work within our group, including rumour stance and veracity classification classification, predicting well-being based on heterogeneous user generated data and target-dependent sentiment recognition. False information circulating on social media presents many risks as social media is used as a source of news by many users. Detecting rumourous content is important to prevent the spread of false information which can affect important decisions and stock markets. Rumour stance classification is considered to be an important step towards rumour verification as claims that attract a lot of scepticism among users are more likely to be proven false later. Therefore performing stance classification well is expected to be useful in debunking false rumours. In our work we classify a set of Twitter posts from conversations discussing rumours as either supporting, denying, questioning or commenting on the underlying rumours. We propose a LSTM-based sequential model that achieves state-of-the-art results on this task through modelling the conversational structure of tweets. The task of automatically assessing well-being using smartphones and online social media is becoming of crucial importance, as an attempt to help individuals self-monitor their mental health state. In the current work, a multiple kernel learning approach is proposed as a mental health predictor, trained on heterogeneous (text and smartphone) user-generated data. The results reveal the efficiency of the proposed model and sequential approaches for time series modelling (i.e., LSTMs) are proposed for future work. Opinion mining is usually achieved by determining the overall sentiment expressed. However, inferring the sentiment towards specific targets is limited by such an approach since a Social Media posts may contain different types of sentiment expressed towards each of the targets mentioned. Our work on target-specific sentiment recognition goes beyond tweet-level or single-target approaches, and proposes a multi-target-specific sentiment classification model, which explores the context around a target as well as syntactic dependencies involving the target.
I am a Computer Science PhD student at the University of Warwick supervised by Dr. Maria Liakata and Prof. Rob Procter. Currently I am based at the Alan Turing Institute in London. My background is Applied mathematics and Complexity science. I work in the area of Natural Language Processing. My research is focused on Rumour Stance and Veracity Classification in Twitter conversations. I am studying the benefits of utilising the conversation structure in supervised learning models.




Noura Al Moubayed - Assistant Professor - Durham University
Applications of DL: Anomaly Detection, Sentiment Classification, & Over-sampling
Noura Al Moubayed - Durham University
Applications of Deep Learning: Anomaly Detection, Sentiment Classification, and Over-sampling
The talk will present my latest work in deep learning for advancing three strands: I) anomaly detection using stacked denoising auto-encoders. II) Sentiment analysis of variable size datasets using a combined LDA and deep learning approach III) deep over-sampling to overcome class bias in highly imbalance datasets. The methods are demonstrated through real-life problems including: spam filtering, sentiment analysis, and Brain-Computer Interfaces.
Dr. Al Moubayed is an Assistant Professor at the school of computer science in Durham University. Her main research interest is in unsupervised deep learning, natural language processing, and optimisation. Dr. Almoubayed obtained her PhD from the Robert Gordon University, followed by post-doctoral positions in the University of Glasgow and Durham University. She developed machine learning and deep learning solutions in the areas of social signal processing, cyber-security, and Brain-Computer Interfaces. All of which involve high dimensional, noisy and imbalance data challenges.




Xavier Giro-i-Nieto - Associate Professor - Universitat Politècnica de Catalunya
One Perceptron to Rule Them All
Xavier Giro-i-Nieto - Universitat Politècnica de Catalunya
Cross-Modal Machine Translation
The advances on neural machine translation across natural language have opened new venues in the field of cross-modal analysis. Given the unified the machine learning framework broadly adopted by the language and vision communities, novel opportunities have arisen by using deep learning framework to transform across modalities. This talk will provide an overview of the state of the art on cross-modal translation and (eg. lipreading, facial animation, sign language) and present our work in speaker visualization from speech.
Xavier Giro-i-Nieto is a learning enthusiast working as an associate professor at the Universitat Politecnica de Catalunya (UPC), in Barcelona, and a certified instructor at the NVIDIA Deep Learning Institute. He has been a visiting scholar at Columbia University and works regularly with Dublin City University, the Barcelona Supercomputing Center and Vilynx. His research interests focus on deep learning for computer vision, speech and natural language processing. His current service includes associate editor of the IEEE Transactions in Multimedia. Xavier Giro-i-Nieto is a learning enthusiast working as an associate professor at the Universitat Politecnica de Catalunya (UPC), in Barcelona, and a certified instructor at the NVIDIA Deep Learning Institute. He has been a visiting scholar at Columbia University and works regularly with Dublin City University, the Barcelona Supercomputing Center and Vilynx. His research interests focus on deep learning for computer vision and natural language processing applied to large scale image retrieval, affective computing, lifelogging from wearables and visual saliency prediction. His current service includes associate editor of the IEEE Transactions in Multimedia and ACM SIGMM Records.



COFFEE
COMPUTER VISION


Antonia Creswell - DL & CV PhD Student - Imperial College London
Iterative Approach To Improving Sample Generation
Antonia Creswell - Imperial College London
Iterative Approach To Improving Sample Generation
Generative modelling allows us to synthesise new data samples, whether the samples be used for imagining new concepts, augmenting datasets or better understanding our datasets. When synthesising samples, generative models don’t always get it right first time - the samples may not be sharp, may have artefacts or may be nonsensical. I will present recent research that shows how we can improve these samples, by applying an iterative procedure to get slightly better samples with each step.
Antonia is a PhD candidate at Imperial College London, in the Bio-Inspired Computer Vision Group. Her research focuses on unsupervised learning and generative models. She received her masters in Biomedical Engineering from Imperial College London with an exchange year at the University of California, Davis. Antonia has interned at DeepMind, Twitter (Magic Pony), Cortexica and UNMADE.




Zeynep Akata - Assistant Professor - University of Amsterdam
Discovering & Synthesizing Novel Concepts With Minimal Supervision
Zeynep Akata - University of Amsterdam
Discovering and synthesizing novel concepts with minimal supervision
Scaling up visual category recognition to large numbers of classes remains challenging. A promising research direction is zero-shot learning, which does not require any training data to recognize new classes, but rather relies on some form of auxiliary information describing new classes. We propose and compare different class embeddings learned automatically from unlabeled text corpora, expert annotated attributes and detailed visual descriptions. Furthermore, we use detailed visual descriptions to generate images from scratch and to generate visual explanations which justify a classification decision. Finally, we explore humans' natural ability to determine distinguishing properties of unknown objects through gaze fixations.
Zeynep Akata is an Assistant Professor at the University of Amsterdam and a Senior Researcher at the Max Planck Institute for Informatics. She received a MSc degree in 2011 from RWTH Aachen and a PhD degree in 2014 from INRIA Grenoble. Her research interests include machine learning with applications to computer vision, such as zero-shot learning and multimodal deep learning with generative models that combine vision and language. She received Lise-Meitner Award for Excellent Women in Computer Science from Max Planck Society in 2014 and a DARPA grant Explainable Artificial Intelligence in 2017 in collaboration with UC Berkeley.

ACCURACY & APPLICATION


Fangde Liu - Research Associate - Imperial College London
Cooperative AI for Critical Application
Fangde Liu - Imperial College London
Cooperative AI for Critical Application
Deep Learning or AI excel at prediction accuracy, but hard to interprete, which becomes a key barrier for many impactful critical applications, such like surgery and finance. Dr Fangde Liu from DSI will talk about their research and idea to make the machine learning system interpretative, controllable, data efficient to train and with fault detection functionality. In his talk , he will talk about the Deep Poincare Map algorithm that combine the dynamic theory with deep learning to perform medical image analysis, and also the autonomous vigiliant management system for one line AI applications, the TensorDB framework.
Dr Fangde Liu, UK National Exceptional Talent, Senior Associate of Royal Society of Medicine, Member of British Machine Vision Society, is an expert in Data Science, Robotics and Surgery. At the Data Science Institute, he manages the engineering development of several big data healthcare projects of global impact. The Optimise-MS, an international joint effort for cloud based autonomous system for effectiveness, performance and post market pharmacovigilance, has ben deployed at about 20 leading clinical institutes across the UK, Belgium, Poland and Finland. As it becomes the technology foundation for neurology registry of several European nations, in the coming decade, it will deployed and monitor over $10 Billion neurology drugs usage yearly across 90 countries. His earlier works on real-time image guided surgery navigation system and robotic operating system on GPU are key technologies of several EU flagship surgical robot projects for endovasc ular surgery, neurosurgery, and orthopaedics, including the EDEN2020, the biggest ever surgical robot project funded in EU. His AI research focus on cooperative AI system aiming breaking down the barriers for critical applications. He focuses on the idea combining domain knowledge and deep learning to make AI system interpretable, controllable and data efficient, he also works on the vigilance management of AI for critical applications. Dr Liu is very experienced and has great insight applying technology in clinical medicine. He works closely with world top clinicians, industry partners and leading academics, such as Johnson&Johnson, Biogen, GE, Renishaw, St Mary Hospital, Mayo Clinic. As an official consultant of Imperial College, he has also provided consulting service on clinical applications of AI to technology giants such as IBM Watson, Huawei Research, NEC research, Samsung Electronics HNA group and AliCloud Europe.




Luca Perletta - Technical Engineer - DigitalGlobe
Artificial Intelligence for Space-Based Technology
Luca Perletta - DigitalGlobe
Artificial Intelligence for Space-Based Technology
Understanding change is essential to addressing our most pressing global challenges. As populations grow, infrastructure must expand. As globalization spreads, businesses must react. As the environment changes, humans must adapt.
To make sense of these developments, we need actionable insight and informed collaboration now more than ever. Today, the ability to observe, analyze and monitor our planet is unprecedented.
DigitalGlobe owns and operates the most sophisticated satellite imaging constellation. This technology has enabled an image library spanning over 25 years and amounting to over 100 Petabytes of high resolution imagery. To be able to process and make sense of such as vast amount of data, at scale, DigitalGlobe is pioneering the use of deep learning and artificial intelligence. From establishing economic indicators, to automatedly detect objects and make predictions about future events, artificial intelligence applied to geospatial information is helping reveal insights that would otherwise be hidden.
Luca has over 5 years experience in developing applications that rely on space-based radar and optical data. Luca currently serves as a Technical Engineer at DigitalGlobe where he is responsible for creating solutions based on satellite imagery, geospatial big data and deep learning algorithms. Academically, Luca holds an MSc in signal processing and remote sensing.



CONVERSATION & DRINKS
ATTENDEE DINNER - County Hall, Westminster Bridge
Attendees and speakers from the Deep Learning Summit & AI Assistant Summit will come together for an evening of discussions, collaborations and merriment after the first day of presentations. Network with your peers and make new contacts. Diners will include CTOs, CEOs, Data Scientists, Founders, Professors and Engineers.
Enjoy a champagne reception with views over the River Thames followed by dinner. Attendees will enjoy extended networking over 3 courses of the finest British cuisine from Chef Alan with wine to match and compliment the evening.
This dinner is only open to attendees of the Deep Learning Summit and AI Assistant Summit. For enquiries about the event please contact [email protected]

DOORS OPEN

WELCOME
Antonia Creswell - Imperial College London
Iterative Approach To Improving Sample Generation
Generative modelling allows us to synthesise new data samples, whether the samples be used for imagining new concepts, augmenting datasets or better understanding our datasets. When synthesising samples, generative models don’t always get it right first time - the samples may not be sharp, may have artefacts or may be nonsensical. I will present recent research that shows how we can improve these samples, by applying an iterative procedure to get slightly better samples with each step.
Antonia is a PhD candidate at Imperial College London, in the Bio-Inspired Computer Vision Group. Her research focuses on unsupervised learning and generative models. She received her masters in Biomedical Engineering from Imperial College London with an exchange year at the University of California, Davis. Antonia has interned at DeepMind, Twitter (Magic Pony), Cortexica and UNMADE.


STARTUP SESSION


Tony Beltramelli - Founder and CEO - uizard.io
Towards Automatic Front-end Development With Deep Learning
Tony Beltramelli - uizard.io
Towards Automatic Front-end Development With Deep Learning
Transforming a graphical user interface mockup created by a designer into computer code is a time-consuming process conducted by a front-end developer in order to build customized software, websites and mobile applications. In this session I will describe how I designed pix2code, a simple deep neural network trained end-to-end to automatically generate computer code given a user interface screenshot as input; finally I will highlight possible future development for the generation of programs from visual input.
Tony is the current CEO and founder of UIzard Technologies. He specialized in machine learning during his graduate studies at the IT University of Copenhagen and ETH Zurich. His research work on the application of deep learning was featured in international media such as WIRED, Forbes, The Huffington Post, The Next Web, Gizmodo and more.




Christopher Bonnett - Senior Machine Learning Researcher - alpha-i
Bayesian DL for Accurate Characterisation of Uncertainties in Time Series Analysis
Christopher Bonnett - alpha-i
Bayesian deep learning for accurate characterisation of uncertainties in time series analysis
At alpha-i we are developing deep learning models for accurate characterisation of uncertainties in time series analysis. We achieve this by combining deep learning methodologies with powerful Bayesian formalism. The alpha-i deep learning network is able not only to make forecasts from time series but also to associate each prediction with a confidence level, which is derived from the information about the model and the data available. One of the key aspect of this Bayesian deep learning methodology is its aversion to over-fitting obtained thanks to the robust probabilistic inference framework. We are also developing novel Bayesian inference methodologies to significantly boost the online performance of our machinery.
Christopher Bonnett has a Masters in Astronomy from the University of Leiden and a PhD in Cosmology from the University of Pierre et Marie Curie. He has 6 years of post-doctoral experience as a key member of several large international collaborations measuring the accelerated expansion of the universe. He has extensive experience in applying deep learning to inverse problems in astronomy. He attended the Insight data science fellowship program in NYC.




Antoine Amann - CEO & Founder - Echobox
Using Deep Learning to Understand the Meaning of Content
Antoine Amann - Echobox
Using Deep Learning to Understand the Meaning of Content
Artificial intelligence is set to revolutionise social media marketing and fundamentally change how content creators interact with data. In a data-abundant world, data analytics has become a $200 billion market. Yet too many analytics dashboards describe data without providing actionable insights, forcing users to rely on gut feeling to translate metrics into actions. AI can distill any amount of data into optimal strategies and even implement those strategies - if humans and machines learn to cooperate. Having brought AI into hundreds of social media teams from London to Lima and from Nairobi to New York, Echobox founder Antoine Amann will speak about the potential of AI and his unique insights into how we can harness the potential of technology.
Antoine is the founder and CEO of Echobox, an AI startup working with the world’s biggest publishers, including The Guardian. Every month, social media posts created by the Echobox AI appear over 10 billion times in a user’s news feed. The long-term Echobox vision is to disrupt the data science industry by creating an AI-first content management tool that users don’t even have to log into - the driverless car of social media management. Echobox also wants to continue to introduce the latest AI technology into Asian, African and Latin American newsrooms. As one of Le Monde’s best-known journalists put it, Antoine’s idea is a “win-win” for newsrooms everywhere. Before building the first version of Echobox from scratch, Antoine got first-hand experience in the world of journalism at the Financial Times in London, and he is passionate about leveraging the potential of artificial intelligence to help publishers to continue to deliver high-quality news in an era of scarce resources.


Ed Newton-Rex - Jukedeck
Ed Newton-Rex is the Founder & CEO of London-based artificial intelligence startup Jukedeck, where he and his team are building AI that can compose original music.
Ed learnt to code in order to start Jukedeck, which now comprises a team of 20 musicians and engineers. Jukedeck has been named one of WIRED’s Hottest European Startups and has won a number of startup competitions, including the Startup Battlefield at TechCrunch Disrupt and LeWeb in Paris, as well as a Cannes Innovation Lion.
Ed spent his childhood studying, performing and composing music, and graduated from Cambridge University with a double-starred First in Music in 2010. He is a published composer, and wrote music for a number of choral groups and for theatre before starting Jukedeck.



COFFEE


Eli David - CTO - Deep Instinct
End-to-End Deep Learning for Detection, Prevention, & Classification of Cyber Attacks
Eli David - Deep Instinct
End-to-End Deep Learning for Detection, Prevention, and Classification of Cyber Attacks
With more than a million new malicious files created every single day, it is becoming exceedingly difficult for currently existing malware detection methods to detect most of these new sophisticated attacks. In this talk, we describe how Deep Instinct uses an end-to-end deep learning based approach to effectively train its brain on hundreds of millions of files, and thus providing by far the highest detection and prevention rates in the cybersecurity industry today. We will additionally explain how deep learning is employed for malware classification and attribution of attacks to specific entities.
Dr. Eli David is a leading expert in the field of computational intelligence, specializing in deep learning (neural networks) and evolutionary computation. He has published more than thirty papers in leading artificial intelligence journals and conferences, mostly focusing on applications of deep learning and genetic algorithms in various real-world domains. For the past ten years, he has been teaching courses on deep learning and evolutionary computation, in addition to supervising the research of graduate students in these fields. He has also served in numerous capacities successfully designing, implementing, and leading deep learning based projects in real-world environments. Dr. David is the developer of Falcon, a grandmaster-level chess playing program based on genetic algorithms and deep learning. The program reached the second place in World Computer Speed Chess Championship. He received the Best Paper Award in 2008 Genetic and Evolutionary Computation Conference, the Gold Award in the prestigious "Humies" Awards for Human-Competitive Results in 2014, and the Best Paper Award in 2016 International Conference on Artificial Neural Networks. Currently Dr. David is the co-founder and CTO of Deep Instinct, the first company to apply deep learning to cybersecurity. Recently Deep Instinct was recognized by Nvidia as the "most disruptive AI startup".


ROBOTICS & DEEP LEARNING


Ankur Handa - Research Scientist - OpenAI
The Quest for Understanding Real World With Synthetic Data
Ankur Handa - OpenAI
The Quest for Understanding Real World With Synthetic Data
Understanding real world involves recognising objects, their physical locations and different relationships among them. This is a much higher level scene understanding than 3D reconstruction and camera pose, and has relied mainly on supervised training data which is laborious to obtain and consequently limited. In this work, we show how photo-realistic simulations and computer graphics can provide the necessary data to ameliorate this problem and help us get a better understanding of real world.
Ankur obtained his PhD in Prof. Andrew Davison's robotvision lab at Imperial College London on real-time SLAM and camera tracking. He finished his post-doctoral research at the University of Cambridge with Prof. Roberto Cipolla on scene understanding with his work on SceneNet. He then returned to Andrew's lab as a Dyson Research Fellow and continued his work on using simulations to generate data for machine learning to do scene understanding with SceneNet RGB-D. He is now with OpenAI as a Research Scientist and works on 3D scene understanding for robotics, RL, and transfer learning.




Jörg Bornschein - Research Scientist - DeepMind
Memory & Rapid Adaption in Generative Models
Jörg Bornschein - DeepMind
Memory & Rapid Adaption in Generative Models
One of the most important algorithms in deep learning is stochastic gradient descent and its variants: slowly adapting the models parameters one mini-batch at a time. But we sometimes face situations where we would like to rapidly adapt our models based on only very few training examples. This situation is called few-shot learning and might arise in supervised, unsupervised and in reinforcement learning. Here I will talk about recent approaches to augment generative models with memory subsystems, how they add few-shot learning capabilities to our models, and how to generate new samples based on very few training examples.
Jorg Bornschein was previously a Global Scholar with the Canadian Institute for Advanced Research (CIFAR) and postdoctoral researcher in Yoshua Bengio’s machine learning lab at the University of Montreal. He is currently concentrating on unsupervised and semisupervised learning using deep architectures. Before moving to Montreal Jorg obtained his PhD from the University of Frankfurt working on large scale bayesian inference for non-linear sparse coding with a focus on building maintainable and massive parallel implementations for HPC clusters. Jorg was also chair and one of the founders of the german hackerspace “Das Labor” which was awarded in 2005 by the federal government for promoting STEM programs to prospective students.


LUNCH


Feryal Behbahani - Postdoc Research Associate - Imperial College London
What Would it Take to Train an Agent to Play with a Shape-Sorter?
Feryal Behbahani - Imperial College London
What Would it Take to Train an Agent to Play with a Shape-Sorter?
The capabilities of humans to precisely and robustly recognise and manipulate objects has been instrumental in the development of human cognition. However, understanding and replicating this process has proven to be difficult. This is of particular importance when thinking of agents or robots acting in naturalistic environments, solving complex tasks. I will present recent work in this direction, focusing on computational optimality and Deep Reinforcement Learning techniques, to discover how to manipulate objects within a 3D physics simulator from high-dimensional sensory observations.
Feryal has received her PhD from the Department of Computing at Imperial College London where she studied Computational Neuroscience and Machine Learning at the Brain and Behaviour Lab. Her main research focused on investigating the underlying algorithms employed by the human brain for object representation and inference. She has previously obtained her MSc in Artificial Intelligence with distinction at Imperial College London. She has also worked on projects building machine learning solutions as part of a technology consultancy start-up that she co-founded. Currently, she is a visiting postdoctoral researcher at Imperial College London where she works on transfer learning and deep reinforcement learning.


APPLICATIONS OF DEEP LEARNING


Maggie Mhanna - Data Scientist - Renault Digital
Artificial Intelligence Driving the Future of Connected Cars
Maggie Mhanna - Renault Digital
Artificial Intelligence Driving the Future of Connected Cars
The emergence of Big Data, machine learning and advanced algorithms to imitate the cognitive functions of the human mind, has begun to simplify and enhance even the simplest aspects of our everyday experiences — and the automotive industry is no exception. Machine learning in the automotive industry has a remarkable ability to bring out hidden relationships among data sets and make predictions, which can lead to an increased level of accuracy in decision-making and improved performance. In this talk, we will be focusing on how machine learning algorithms can aid in effective planning and execution of predictive maintenance; predicting what is likely to fail and when it is going to happen. How it can accurately incorporate analysis results of customer feedback which helps in building vehicle and sub-systems performance for guiding future product design, and so forth.
Maggie Mhanna is a Data Scientist at Renault Digital, and a part-time university professor Leonardo da Vinci Engineering School. Her work now focuses on the application of data science and machine learning in the connected cars industry. Before joining Renault Digital, Maggie was doing a PhD at Centrale-Supélec in France allowing her to publish various articles in the area of machine learning, signal processing and information theory. Her thesis topic was entitled : "Privacy-Preserving Quantization Learning for Distributed Detection & Estimation with Applications to Smart Meters". Maggie earned a masters of science in renewable energies from Ecole Polytechnique, and an engineering degree in computer and communication from the Lebanese University.




Timor Kadir - Founder & CSTO - Optellum
Balance, Bias & Size: Unlocking the Potential of Retrospective Clinical Data for Healthcare Applications
Timor Kadir - Optellum
Balance, bias and size: unlocking the potential of retrospective clinical data for healthcare applications
Since the widespread adoption of IT systems such as PACS (Picture Archival and Communication System) and EMR (Electronic Medical Record), hospitals have been collecting digital patient data as part of clinical practice. Such retrospective archives represent a treasure trove for deep learning applications in healthcare. However, in practice such data is rarely in a form that can be readily utilized for learning due to missing labels, class imbalance and bias. In this talk, I will discuss techniques to address such challenges including self-supervision, multi-task learning and transfer learning and will illustrate using applications in lung cancer and chronic backpain.
Timor Kadir graduated with an MEng in Electrical and Electronic Engineering from Surrey University in 1996 and studied for a DPhil at University of Oxford under Sir Michael Brady. He joined CTI/Siemens as a Research Scientist working on computer vision software and in 2009, during a management buy-out, Timor became CTO for Mirada Medical. Most recently, he founded Optellum, a company delivering machine learning based clinical risk stratification. He is currently the CTO of Optellum, Mirada and visiting fellow at Oxford. He’s published a reasonable number of papers, filed a bunch of patents and his h-index isn’t bad either.



Noor Shaker - Co-Founder & CEO - GTN
Drug Discovery Disrupted: Quantum Physics Meets Machine Learning
Noor Shaker - GTN
Drug Discovery Disrupted: Quantum Physics Meets Machine Learning
Whenever a disease is identified, a new journey into the “chemical space” starts seeking a medicine that could become useful in contending diseases. The journey takes approximately 15 years and costs $2.6bn, and starts with a process to filter millions of molecules to identify the promising hundreds with high potential to become medicines. Around 99% of selected leads fail later in the process due to inaccurate prediction of behaviour and the limited pool from which they were sampled. We are at GTN Ltd addressing the main bottlenecks in drug development by new innovations marring ideas from machine learning and quantum physics.
Prof. Noor Shaker is a co-founder and CEO at GTN Ltd. Before starting GTN, she was an assistant professor at Aalborg University in Copenhagen working on different aspects of machine learning with special interest in generative models. She is the main author of the book “Procedural Content Generation in Games” which covers many of the generative methods. She has more than 50 publications and 1000+ citations. She serves as the chair of the IEEE games technical committee and she is an active member in the games research society participating in organizing conferences, workshops and tasks forces. She has won a number of awards for her research including the IEEE Transactions on Computational Intelligence and AI in Games Outstanding Paper Award. At GTN, she is working with leading researchers on a novel, patent-pending, technology to drug discovery bringing ideas from quantum physics and machine learning.



END OF SUMMIT