Cambridge, United Kingdom

Time filter

Source Type

Demaine E.D.,Cambridge Intelligence | Schulz A.,University of Hagen
Discrete and Computational Geometry | Year: 2017

A stacking operation adds a d-simplex on top of a facet of a simplicial d-polytope while maintaining the convexity of the polytope. A stacked d-polytope is a polytope that is obtained from a d-simplex and a series of stacking operations. We show that for a fixed d every stacked d-polytope with n vertices can be realized with nonnegative integer coordinates. The coordinates are bounded by (Formula presented.), except for one axis, where the coordinates are bounded by (Formula presented.). The described realization can be computed with an easy algorithm. The realization of the polytopes is obtained with a lifting technique which produces an embedding on a large grid. We establish a rounding scheme that places the vertices on a sparser grid, while maintaining the convexity of the embedding. © 2017 Springer Science+Business Media New York

Popa R.A.,Cambridge Intelligence | Redfield C.M.S.,Cambridge Intelligence | Zeldovich N.,Cambridge Intelligence | Balakrishnan H.,Cambridge Intelligence
Communications of the ACM | Year: 2012

CryptDB is a practical system that explores an intermediate design point to provide confidentiality for applications that use database management systems (DBMSes). CryptDB is the first practical system that can execute a wide range of SQL queries over encrypted data. CryptDB requires no changes to the internals of the DBMS server, and should work with most standard SQL DBMSes. CryptDB's architecture consisting of two parts, a proxy and an unmodified DBMS. It uses user-defined functions (UDFs) to perform cryptographic operations in the DBMS. The only information that CryptDB reveals to the DBMS server is relationships among data items corresponding to classes of computation that queries perform on the database, such as comparing items for equality, sorting, or performing word search. RND provides the maximum security in CryptDB, indistinguishability under an adaptive chosen plain text attack (IND-CPA).

Benczur A.A.,Hungarian Academy of Sciences | Karger D.R.,Cambridge Intelligence
SIAM Journal on Computing | Year: 2015

We describe random sampling techniques for approximately solving problems that involve cuts and flows in graphs. We give a near-linear-time randomized combinatorial construction that transforms any graph on n vertices into an O(n log n)-edge graph on the same vertices whose cuts have approximately the same value as the original graph's. In this new graph, for example, we can run the Õ (m3/2)-time maximum flow algorithm of Goldberg and Rao to find an s-t minimum cut in Õ(n3/2) time. This corresponds to a (1 + ε)-times minimum s-t cut in the original graph. A related approach leads to a randomized divide-and-conquer algorithm producing an approximately maximum flow in Õ(m √n) time. Our algorithm can also be used to improve the running time of sparsest cut approximation algorithms from Õ(mn) to Õ (n2) and to accelerate several other recent cut and flow algorithms. Our algorithms are based on a general theorem analyzing the concentration of random graphs' cut values near their expectations. Our work draws only on elementary probability and graph theory. © 2015 Andras Benczúr and David R. Karger.

Shankar S.,Cambridge Intelligence | Garg V.K.,Massachusetts Institute of Technology | Cipolla R.,Cambridge Intelligence
Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition | Year: 2015

Most of the approaches for discovering visual attributes in images demand significant supervision, which is cumbersome to obtain. In this paper, we aim to discover visual attributes in a weakly supervised setting that is commonly encountered with contemporary image search engines. For instance, given a noun (say forest) and its associated attributes (say dense, sunlit, autumn), search engines can now generate many valid images for any attribute-noun pair (dense forests, autumn forests, etc). However, images for an attributenoun pair do not contain any information about other attributes (like which forests in the autumn are dense too). Thus, a weakly supervised scenario occurs: each of the M attributes corresponds to a class such that a training image in class m ε {1, . . . ,M} contains a single label that indicates the presence of the mth attribute only. The task is to discover all the attributes present in a test image. Deep Convolutional Neural Networks (CNNs) [20] have enjoyed remarkable success in vision applications recently. However, in a weakly supervised scenario, widely used CNN training procedures do not learn a robust model for predicting multiple attribute labels simultaneously. The primary reason is that the attributes highly co-occur within the training data, and unlike objects, do not generally exist as well-defined spatial boundaries within the image. To ameliorate this limitation, we propose Deep-Carving, a novel training procedure with CNNs, that helps the net efficiently carve itself for the task of multiple attribute prediction. During training, the responses of the feature maps are exploited in an ingenious way to provide the net with multiple pseudo-labels (for training images) for subsequent iterations. The process is repeated periodically after a fixed number of iterations, and enables the net carve itself iteratively for efficiently disentangling features. Additionally, we contribute a noun-adjective pairing inspired Natural Scenes Attributes Dataset to the research community, CAMIT - NSAD, containing a number of co-occurring attributes within a noun category. We describe, in detail, salient aspects of this dataset. Our experiments on CAMITNSAD and the SUN Attributes Dataset [29], with weak supervision, clearly demonstrate that the Deep-Carved CNNs consistently achieve considerable improvement in the precision of attribute prediction over popular baseline methods. © 2015 IEEE.

Loh P.-R.,Harvard University | Loh P.-R.,The Broad Institute of MIT and Harvard | Tucker G.,Harvard University | Tucker G.,Massachusetts Institute of Technology | And 17 more authors.
Nature Genetics | Year: 2015

Linear mixed models are a powerful statistical tool for identifying genetic associations and avoiding confounding. However, existing methods are computationally intractable in large cohorts and may not optimize power. All existing methods require time cost O(MN 2) (where N is the number of samples and M is the number of SNPs) and implicitly assume an infinitesimal genetic architecture in which effect sizes are normally distributed, which can limit power. Here we present a far more efficient mixed-model association method, BOLT-LMM, which requires only a small number of O(MN) time iterations and increases power by modeling more realistic, non-infinitesimal genetic architectures via a Bayesian mixture prior on marker effect sizes. We applied BOLT-LMM to 9 quantitative traits in 23,294 samples from the Women's Genome Health Study (WGHS) and observed significant increases in power, consistent with simulations. Theory and simulations show that the boost in power increases with cohort size, making BOLT-LMM appealing for genome-wide association studies in large cohorts. © 2015 Nature America, Inc. All rights reserved.

Che A.J.,Cambridge Intelligence | Che A.J.,Ginkgo BioWorks, Inc. | Knight Jr. T.F.,Cambridge Intelligence | Knight Jr. T.F.,Ginkgo BioWorks, Inc.
Nucleic Acids Research | Year: 2010

Controlling RNA splicing opens up possibilities for the synthetic biologist. The Tetrahymena ribozyme is a model group I self-splicing ribozyme that has been shown to be useful in synthetic circuits. To create additional splicing ribozymes that can function in synthetic circuits, we generated synthetic ribozyme variants by rationally mutating the Tetrahymena ribozyme. We present an alignment visualization for the ribozyme termed as structure information diagram that is similar to a sequence logo but with alignment data mapped on to secondary structure information. Using the alignment data and known biochemical information about the Tetrahymena ribozyme, we designed synthetic ribozymes with different primary sequences without altering the secondary structure. One synthetic ribozyme with 110 nt mutated retained 12% splicing efficiency in vivo. The results indicate that our biochemical understanding of the ribozyme is accurate enough to engineer a family of active splicing ribozymes with similar secondary structure but different primary sequences. © The Author(s) 2010. Published by Oxford University Press.

Agency: GTR | Branch: Innovate UK | Program: | Phase: Innovation Voucher | Award Amount: 5.00K | Year: 2015

We are looking for cyber security expertise to review and improve our current procedures.

News Article | March 20, 2014

KeyLines is a toolkit for building your own data visualization capability and integrating it into your own web applications with very little effort. KeyLines is compatible with any data store, so How you collect and store the data is entirely up to you. The application you build will also run on any device and in any web browser, meaning it can be deployed to anyone who needs to use it. Your data stays in your control at all times: KeyLines is self-contained and needs no external connections.

News Article | July 23, 2014

Cambridge Intelligence, today announced the release of KeyLines 2.0 – the next generation of network visualisation software. This latest release of their web software development kit (or SDK) makes it the first in the world to offer full support for dynamic graphs and temporal visualisation – meaning users can now watch their networks evolving through time in any web browser. This new dimension to data analysis is provided via the KeyLines Time Bar. Using the time bar, data analysts can instantly understand network trends and focus on the points in time that matter most. It is expected to reduce data processing and investigation time significantly, helping analysts to follow trends and predict future events with even greater accuracy. Global organisations, including government agencies, police forces, and companies already rely upon KeyLines to uncover relationships and patterns in large and complex datasets. Many of Cambridge Intelligence’s existing customers are planning to stay ahead of the innovation curve by deploying the time bar to their users this year. Cambridge Intelligence CEO, Joe Parry, said of the new release, “The ability to view networks change through time ensures that KeyLines remains at the forefront of the industry. These new capabilities have generated a significant buzz among current customers and are sure to drive our growth into 2015.” This release comes after a busy six months of success for the three-year-old Cambridge company. In March they celebrated moving into modern new park-side offices on Regent Street, and have enjoyed recognition in innovation and technology awards – most recently being shortlisted for PwC’s Private Business of the Year.

Loading Cambridge Intelligence collaborators
Loading Cambridge Intelligence collaborators