Training-free probability models for whole-image based place recognition

Whole-image descriptors such as GIST have been used successfully for persistent place recognition when combined with temporal filtering or sequential filtering techniques. However, whole-image descriptor localization systems often apply a heuristic rather than a probabilistic approach to place recognition, requiring substantial environmental-specific tuning prior to deployment. In this paper we present a novel online solution that uses statistical approaches to calculate place recognition likelihoods for whole-image descriptors, without requiring either environmental tuning or pre-training. Using a real world benchmark dataset, we show that this method creates distributions appropriate to a specific environment in an online manner. Our method performs comparably to FAB-MAP in raw place recognition performance, and integrates into a state of the art probabilistic mapping system to provide superior performance to whole-image methods that are not based on true probability distributions. The method provides a principled means for combining the powerful change-invariant properties of whole-image descriptors with probabilistic back-end mapping systems without the need for prior training or system tuning.

Do prediction markets produce well-calibrated probability forecasts?

This article presents new theoretical and empirical evidence on the forecasting ability of prediction markets. We develop a model that predicts that the time until expiration of a prediction market should negatively affect the accuracy of prices as a forecasting tool in the direction of a ‘favourite/longshot bias’. That is, high-likelihood events are underpriced, and low-likelihood events are over-priced. We confirm this result using a large data set of prediction market transaction prices. Prediction markets are reasonably well calibrated when time to expiration is relatively short, but prices are significantly biased for events farther in the future. When time value of money is considered, the miscalibration can be exploited to earn excess returns only when the trader has a relatively low discount rate.

Using the quantum probability ranking principle to rank interdependent documents

A known limitation of the Probability Ranking Principle (PRP) is that it does not cater for dependence between documents. Recently, the Quantum Probability Ranking Principle (QPRP) has been proposed, which implicitly captures dependencies between documents through “quantum interference”. This paper explores whether this new ranking principle leads to improved performance for subtopic retrieval, where novelty and diversity is required. In a thorough empirical investigation, models based on the PRP, as well as other recently proposed ranking strategies for subtopic retrieval (i.e. Maximal Marginal Relevance (MMR) and Portfolio Theory(PT)), are compared against the QPRP. On the given task, it is shown that the QPRP outperforms these other ranking strategies. And unlike MMR and PT, one of the main advantages of the QPRP is that no parameter estimation/tuning is required; making the QPRP both simple and effective. This research demonstrates that the application of quantum theory to problems within information retrieval can lead to significant improvements.

Developing the quantum probability ranking principle

In this work, we summarise the development of a ranking principle based on quantum probability theory, called the Quantum Probability Ranking Principle (QPRP), and we also provide an overview of the initial experiments performed employing the QPRP. The main difference between the QPRP and the classic Probability Ranking Principle, is that the QPRP implicitly captures the dependencies between documents by means of quantum interference". Subsequently, the optimal ranking of documents is not based solely on documents' probability of relevance but also on the interference with the previously ranked documents. Our research shows that the application of quantum theory to problems within information retrieval can lead to consistently better retrieval effectiveness, while still being simple, elegant and tractable.

Integrated approach for diagnostics and prognostics of HP LNG pump based on health state probability estimation

Effective machine fault prognostic technologies can lead to elimination of unscheduled downtime and increase machine useful life and consequently lead to reduction of maintenance costs as well as prevention of human casualties in real engineering asset management. This paper presents a technique for accurate assessment of the remnant life of machines based on health state probability estimation technique and historical failure knowledge embedded in the closed loop diagnostic and prognostic system. To estimate a discrete machine degradation state which can represent the complex nature of machine degradation effectively, the proposed prognostic model employed a classification algorithm which can use a number of damage sensitive features compared to conventional time series analysis techniques for accurate long-term prediction. To validate the feasibility of the proposed model, the five different level data of typical four faults from High Pressure Liquefied Natural Gas (HP-LNG) pumps were used for the comparison of intelligent diagnostic test using five different classification algorithms. In addition, two sets of impeller-rub data were analysed and employed to predict the remnant life of pump based on estimation of health state probability using the Support Vector Machine (SVM) classifier. The results obtained were very encouraging and showed that the proposed prognostics system has the potential to be used as an estimation tool for machine remnant life prediction in real life industrial applications.

Probability and story-telling : normative and descriptive models of juridical proof

The operation of the law rests on the selection of an account of the facts. Whether this involves prediction or postdiction, it is not possible to achieve certainty. Any attempt to model the operation of the law completely will therefore raise questions of how to model the process of proof. In the selection of a model a crucial question will be whether the model is to be used normatively or descriptively. Focussing on postdiction, this paper presents and contrasts the mathematical model with the story model. The former carries the normative stamp of scientific approval, whereas the latter has been developed by experimental psychologists to describe how humans reason. Neil Cohen's attempt to use a mathematical model descriptively provides an illustration of the dangers in not clearly setting this parameter of the modelling process. It should be kept in mind that the labels 'normative' and 'descriptive' are not eternal. The mathematical model has its normative limits, beyond which we may need to critically assess models with descriptive origins.

The Quantum Probability Ranking Principle for Information Retrieval

While the Probability Ranking Principle for Information Retrieval provides the basis for formal models, it makes a very strong assumption regarding the dependence between documents. However, it has been observed that in real situations this assumption does not always hold. In this paper we propose a reformulation of the Probability Ranking Principle based on quantum theory. Quantum probability theory naturally includes interference effects between events. We posit that this interference captures the dependency between the judgement of document relevance. The outcome is a more sophisticated principle, the Quantum Probability Ranking Principle, that provides a more sensitive ranking which caters for interference/dependence between documents’ relevance.

Polarization of forecast densities : a new approach to time series classification

Time series classification has been extensively explored in many fields of study. Most methods are based on the historical or current information extracted from data. However, if interest is in a specific future time period, methods that directly relate to forecasts of time series are much more appropriate. An approach to time series classification is proposed based on a polarization measure of forecast densities of time series. By fitting autoregressive models, forecast replicates of each time series are obtained via the bias-corrected bootstrap, and a stationarity correction is considered when necessary. Kernel estimators are then employed to approximate forecast densities, and discrepancies of forecast densities of pairs of time series are estimated by a polarization measure, which evaluates the extent to which two densities overlap. Following the distributional properties of the polarization measure, a discriminant rule and a clustering method are proposed to conduct the supervised and unsupervised classification, respectively. The proposed methodology is applied to both simulated and real data sets, and the results show desirable properties.

A large proportion of asymptomatic Plasmodium infections with low and sub-microscopic parasite densities in the low transmission setting of Temotu Province, Solomon Islands : challenges for malaria diagnostics in an elimination setting

Background Many countries are scaling up malaria interventions towards elimination. This transition changes demands on malaria diagnostics from diagnosing ill patients to detecting parasites in all carriers including asymptomatic infections and infections with low parasite densities. Detection methods suitable to local malaria epidemiology must be selected prior to transitioning a malaria control programme to elimination. A baseline malaria survey conducted in Temotu Province, Solomon Islands in late 2008, as the first step in a provincial malaria elimination programme, provided malaria epidemiology data and an opportunity to assess how well different diagnostic methods performed in this setting. Methods During the survey, 9,491 blood samples were collected and examined by microscopy for Plasmodium species and density, with a subset also examined by polymerase chain reaction (PCR) and rapid diagnostic tests (RDTs). The performances of these diagnostic methods were compared. Results A total of 256 samples were positive by microscopy, giving a point prevalence of 2.7%. The species distribution was 17.5% Plasmodium falciparum and 82.4% Plasmodium vivax. In this low transmission setting, only 17.8% of the P. falciparum and 2.9% of P. vivax infected subjects were febrile (≥38°C) at the time of the survey. A significant proportion of infections detected by microscopy, 40% and 65.6% for P. falciparum and P. vivax respectively, had parasite density below 100/μL. There was an age correlation for the proportion of parasite density below 100/μL for P. vivax infections, but not for P. falciparum infections. PCR detected substantially more infections than microscopy (point prevalence of 8.71%), indicating a large number of subjects had sub-microscopic parasitemia. The concordance between PCR and microscopy in detecting single species was greater for P. vivax (135/162) compared to P. falciparum (36/118). The malaria RDT detected the 12 microscopy and PCR positive P. falciparum, but failed to detect 12/13 microscopy and PCR positive P. vivax infections. Conclusion Asymptomatic malaria infections and infections with low and sub-microscopic parasite densities are highly prevalent in Temotu province where malaria transmission is low. This presents a challenge for elimination since the large proportion of the parasite reservoir will not be detected by standard active and passive case detection. Therefore effective mass screening and treatment campaigns will most likely need more sensitive assays such as a field deployable molecular based assay.

Development of fourth-grade students’ understanding of experimental and theoretical probability

Students explored variation and expectation in a probability activity at the end of the first year of a 3-year longitudinal study across grades 4-6. The activity involved experiments in tossing coins both manually and with simulation using the graphing software, TinkerPlots. Initial responses indicated that the students were aware of uncertainty, although an understanding of chance concepts appeared limited. Predicting outcomes of 10 tosses reflected an intuitive notion of equiprobability, with little awareness of variation. Understanding the relationship between experimental and theoretical probability did not emerge until multiple outcomes and representations were generated with the software.

Cost-effective suppression and eradication of invasive predators

Introduced predators can have pronounced effects on naïve prey species; thus, predator control is often essential for conservation of threatened native species. Complete eradication of the predator, although desirable, may be elusive in budget-limited situations, whereas predator suppression is more feasible and may still achieve conservation goals. We used a stochastic predator-prey model based on a Lotka-Volterra system to investigate the cost-effectiveness of predator control to achieve prey conservation. We compared five control strategies: immediate eradication, removal of a constant number of predators (fixed-number control), removal of a constant proportion of predators (fixed-rate control), removal of predators that exceed a predetermined threshold (upper-trigger harvest), and removal of predators whenever their population falls below a lower predetermined threshold (lower-trigger harvest). We looked at the performance of these strategies when managers could always remove the full number of predators targeted by each strategy, subject to budget availability. Under this assumption immediate eradication reduced the threat to the prey population the most. We then examined the effect of reduced management success in meeting removal targets, assuming removal is more difficult at low predator densities. In this case there was a pronounced reduction in performance of the immediate eradication, fixed-number, and lower-trigger strategies. Although immediate eradication still yielded the highest expected minimum prey population size, upper-trigger harvest yielded the lowest probability of prey extinction and the greatest return on investment (as measured by improvement in expected minimum population size per amount spent). Upper-trigger harvest was relatively successful because it operated when predator density was highest, which is when predator removal targets can be more easily met and the effect of predators on the prey is most damaging. This suggests that controlling predators only when they are most abundant is the "best" strategy when financial resources are limited and eradication is unlikely. © 2008 Society for Conservation Biology.

Low probability differentials and the cryptanalysis of full-round CLEFIA-128

So far, low probability differentials for the key schedule of block ciphers have been used as a straightforward proof of security against related-key differential analysis. To achieve resistance, it is believed that for cipher with k-bit key it suffices the upper bound on the probability to be 2− k . Surprisingly, we show that this reasonable assumption is incorrect, and the probability should be (much) lower than 2− k . Our counter example is a related-key differential analysis of the well established block cipher CLEFIA-128. We show that although the key schedule of CLEFIA-128 prevents differentials with a probability higher than 2− 128, the linear part of the key schedule that produces the round keys, and the Feistel structure of the cipher, allow to exploit particularly chosen differentials with a probability as low as 2− 128. CLEFIA-128 has 214 such differentials, which translate to 214 pairs of weak keys. The probability of each differential is too low, but the weak keys have a special structure which allows with a divide-and-conquer approach to gain an advantage of 27 over generic analysis. We exploit the advantage and give a membership test for the weak-key class and provide analysis of the hashing modes. The proposed analysis has been tested with computer experiments on small-scale variants of CLEFIA-128. Our results do not threaten the practical use of CLEFIA.

In search of good probability assessors: An experimental comparison of elicitation rules for confidence judgments

In this paper, we use an experimental design to compare the performance of elicitation rules for subjective beliefs. Contrary to previous works in which elicited beliefs are compared to an objective benchmark, we consider a purely subjective belief framework (confidence in one’s own performance in a cognitive task and a perceptual task). The performance of different elicitation rules is assessed according to the accuracy of stated beliefs in predicting success. We measure this accuracy using two main factors: calibration and discrimination. For each of them, we propose two statistical indexes and we compare the rules’ performances for each measurement. The matching probability method provides more accurate beliefs in terms of discrimination, while the quadratic scoring rule reduces overconfidence and the free rule, a simple rule with no incentives, which succeeds in eliciting accurate beliefs. Nevertheless, the matching probability appears to be the best mechanism for eliciting beliefs due to its performances in terms of calibration and discrimination, but also its ability to elicit consistent beliefs across measures and across tasks, as well as its empirical and theoretical properties.

On the probability that the determinant of an nxn matrix over a finite field vanishe

An expression is derived for the probability that the determinant of an n x n matrix over a finite field vanishes; from this it is deduced that for a fixed field this probability tends to 1 as n tends to.