Eric McCreath

Inducing shortcuts on a mobile phone interface

Due to size restrictions, mobile phone user interfaces are often difficult to use[8]. In this short paper, we investigated inducing shortcuts to replace the sequence of actions required to complete common tasks on a mobile phone. In particular, we used mobile phone interaction data to evaluate several methods for inducing shortcuts. We considered the balance between maximising interface efficiency and shortcuts that remained stable and hence predictable.

Use of SIMD Vector Operations to Accelerate Application Code Performance on Low-Powered ARM and Intel Platforms

Augmenting a processor with special hardware that is able to apply a Single Instruction to Multiple Data(SIMD) at the same time is a cost effective way of improving processor performance. It also offers a means of improving the ratio of processor performance to power usage due to reduced and more effective data movement and intrinsically lower instruction counts. This paper considers and compares the NEON SIMD instruction set used on the ARM Cortex-A series of RISC processors with the SSE2 SIMD instruction set found on Intel platforms within the context of the Open Computer Vision (OpenCV) library. The performance obtained using compiler auto-vectorization is compared with that achieved using hand-tuning across a range of five different benchmarks and ten different hardware platforms. On the ARM platforms the hand-tuned NEON benchmarks were between 1.05× and 13.88× faster than the auto-vectorized code, while for the Intel platforms the hand-tuned SSE benchmarks were between 1.34× and 5.54× faster.

An intelligent interface for sorting electronic mail

Classification of email is an important everyday task for a large and growing number of users. This paper describes the i-ems (Intelligent-Electronic Mail Sorter) mail interface, which offers a view of the inbox based on predicted classifications of messages. The interface is designed to ensure user control over the prediction processes by supporting scrutiny of the systems certainty and details of the mechanisms used.

LIME: A System for Learning Relations

This paper describes the design of the inductive logic programming system Lime. Instead of employing a greedy covering approach to constructing clauses, Lime employs a Bayesian heuristic to evaluate logic programs as hypotheses. The notion of a simple clause is introduced. These sets of literals may be viewed as subparts of clauses that are effectively independent in terms of variables used. Instead of growing a clause one literal at a time, Lime efficiently combines simple clauses to construct a set of gainful candidate clauses. Subsets of these candidate clauses are evaluated via the Bayesian heuristic to find the final hypothesis. Details of the algorithms and data structures of Lime are discussed. Limes handling of recursive logic programs is also described. Experimental results to illustrate how LIME achieves its design goals of better noise handling, learning from fixed set of examples (and from only positive data), and of learning recursive logic programs are provided. Experimental results comparing Lime with FOIL and PROGOL in the KRK domain in the presence of noise are presented. It is also shown that the already good noise handling performance of Lime further improves when learning recursive definitions in the presence of noise.

Dynamic Algorithm Selection Using Reinforcement Learning

It is often the case that many algorithms exist to solve a single problem, each possessing different performance characteristics. The usual approach in this situation is to manually select the algorithm which has the best average performance. However, this strategy has drawbacks in cases where the optimal algorithm changes during an invocation of the program, in response to changes in the programs state and the computational environment. This paper presents a prototype tool that uses reinforcement learning to guide algorithm selection at runtime, matching the algorithm used to the current state of the computation. The tool is applied to a simulation similar to those used in some computational chemistry problems. It is shown that the low dimensionality of the problem enables the optimal choice of algorithm to be determined quickly, and that the learning system can react rapidly to phase changes in the target program

SHYSTER-MYCIN: a hybrid legal expert system

SHYSTER-MYCIN combines a case-based legal expert system (SHYSTER) with a rule-based expert system (MYCIN) to form a hybrid legal expert system. MYCINs reporting has been improved for use with SHYSTER-MYCIN to provide more useful information about the systems conclusions.SHYSTER-MYCINs output was tested against that of a group of lawyers, not expert in the test domain (Australian copyright law). This allowed the systems reasoning, rather than its depth of knowledge, to be tested. Testing indicates that SHYSTER-MYCINs approach to the law---using a rule-based system to reason with legislation and a case-based system to reason with cases---is appropriate.

Iems: helping users manage email

This paper reports our work to buildan email interface which can learn how to predict a users email classifications at the same time as ensuring user control over the process. We report our exploration to answer the question: does the classifier work well enough to be effective? There has been considerable work to automate classification of email. Yet, it does not give a good sense of how well we are able to model users classification of email. This paper reports the results of our own evaluations, including a stark observation that evaluation of this class of adaptive system needs to take account of the fact that the user can be expected to adapt to the system. This is important for the long term evaluation of such systems since we may find that this effect means that our systems may be performing better than classic evaluations might suggest.

Performance Evaluation of the NVIDIA GeForce 8800 GTX GPU for Machine Learning

NVIDIA have released a new platform (CUDA) for general purpose computing on their graphical processing units (GPU). This paper evaluates use of this platform for statistical machine learning applications. The transfer rates to and from the GPU are measured, as is the performance of matrix vector operations on the GPU. An implementation of a sparse matrix vector product on the GPU is outlined and evaluated. Performance comparisons are made with the host processor.

Improving the Learning Rate by Inducing a Transition Model

In general, a reinforcement learning agent requires many trials in order to find a successful policy in a domain. In this paper we investigate inducing a transition model to reduce the number of trials required by an agent.We discuss an approach that incorporates transition model learning within a contemporary agent design.

Machine learning for readability of legislative sentences

Improving the readability of legislation is an important and unresolved problem. Recently, researchers have begun to apply legal informatics to this problem. This paper applies machine learning to predict the readability of sentences from legislation and regulations. A corpus of sentences from the United States Code and US Code of Federal Regulations was created. Each sentence was labelled for language difficulty using results from a large-scale crowdsourced study undertaken during 2014. The corpus was used as training and test data for machine learning. The corpus includes a version tagged using the Stanford parser context free grammar and a version tagged using the Stanford dependency grammar parser. The corpus is described and made available to interested researchers. We investigated whether extending natural language features available as input to machine learning improves the accuracy of prediction. Among features evaluated are those from the context free and dependency grammars. Letter and word ngrams were also studied. We found the addition of such features improves accuracy of prediction on legal language. We also undertake a correlation study of natural language features and language difficulty drawing insights as to the characteristics that may make legal language more difficult. These insights, and those from machine learning, enable us to describe a system for reducing legal language difficulty and to identify a number of suggested heuristics for improving the writing of legislation and regulations.