A study on the accuracy of frequency measures and its impact on knowledge discovery in single sequences

In knowledge discovery in single sequences, different results could be discovered from the same sequence when different frequency measures are adopted. It is natural to raise such questions as (1) do these frequency measures reflect actual frequencies accurately? (2) what impacts do frequency measures have on discovered knowledge? (3) are discovered results accurate and reliable? and (4) which measures are appropriate for reflecting frequencies accurately? In this paper, taking three major factors (anti-monotonicity, maximum-frequency and window-width restriction) into account, we identify inaccuracies inherent in seven existing frequency measures, and investigate their impacts on the soundness and completeness of two kinds of knowledge, frequent episodes and episode rules, discovered from single sequences. In order to obtain more accurate frequencies and knowledge, we provide three recommendations for defining appropriate frequency measures. Following the recommendations, we introduce a more appropriate frequency measure. Empirical evaluation reveals the inaccuracies and verifies our findings. 

The accuracy of housing forecasting in Australia

Forecasting is an integral part of all business planning, and forecasting the outlook for housing is of interest to many firms in the housing construction sector. This research measures the performance of a number of industry forecasting bodies; this is done to provide users with an indicator of the value of housing forecasting undertaken in Australia. The accuracy of housing commencement forecasts of three Australian organisations – the Housing Industry Association (HIA), the Indicative Planning Council for the Housing Industry (IPC) and BIS-Shrapnel – is examined through the empirical analysis of their published forecasts supplemented by qualitative data in the form of opinions elicited from several industry “experts” employed in these organisations. Forecasting performance was determined by comparing the housing commencement forecast with the actual data collected by the Australian Bureau of Statistics on an ex-post basis. Although the forecasts cover different time periods, the level of accuracy is similar, at around 11-13 per cent for four-quarter-ahead forecasts. In addition, national forecasts are more accurate than forecasts for individual states. This is the first research that has investigated the accuracy of both private and public sector forecasting of housing construction in Australia. This allows users of the information to better understand the performance of various forecasting organisations.

Accuracy of national mortality codes in identifying adjudicated cardiovascular deaths

Objective: This study investigated the sensitivity and specificity of the national mortality codes in identifying cardiovascular disease (CVD) deaths and documents methods of verification.

Methods: A 12-year retrospective case ascertainment of all ICD-coded CVD deaths was performed for deaths between 1990 and 2002 in the Melbourne Collaborative Cohort Study, comprising 41,528 subjects. Categories of non-CVD codes were also examined. Stratified samples of 750 deaths were adjudicated from a total of 2,230 deaths. Expert panels of cardiologists and neurologists adjudicated deaths.

Results: Of the 750 deaths adjudicated, 582 were verified as CVD [392 coronary heart disease (CHD) and 92 stroke] and 168 non-CVD. Estimated sensitivity and specificity of national mortality codes for identifying specific causes of death were: CHD 74.2% (95% CI: 69.8–78.5%) and 97.6% (96.0–99.2%), respectively; myocardial infarction 59.9% (50.9–69.0%) and 94.2% (92.4–96.0%), respectively; haemorrhagic stroke 58.9% (46.0–71.7%) and 99.8% (99.4–100.0%), respectively and; ischaemic stroke 38.7% (20.5–56.9%) and 99.9% (99.6–100.0%), respectively. Misclassification was most common for deaths with primary ICD codes for endocrine-metabolic and genito-urinary diseases.

Conclusions: National mortality coding under-estimated the true proportion of CHD and stroke deaths in the cohort by 13.6% and 50.8%, respectively.

Retrieval of episodic versus generic information : does the order of recall affect the amount and accuracy of details reported by children about repeated events?

Children (N = 157) 4 to 8 years old participated 1 time (single) or 4 times (repeated) in an interactive event. Across each condition, half were questioned a week later about the only or a specific occurrence of the event (depth first) and then about what usually happens. Half were prompted in the reverse order (breadth first). Children with repeated experience who first were asked about what usually happens reported more eventrelated information overall than those asked about an occurrence first. All children used episodic language when describing an occurrence; however, children with repeated-event experience used episodic language less often when describing what usually happens than did those with a single experience. Accuracy rates did not differ between conditions. Implications for theories of repeated-event memory are discussed.

Predictive validity and classification accuracy of ActiGraph energy expenditure equations and cut-points in young children

Objectives
Evaluate the predictive validity of ActiGraph energy expenditure equations and the classification accuracy of physical activity intensity cut-points in preschoolers.

Methods
Forty children aged 4–6 years (5.3±1.0 years) completed a ~150-min room calorimeter protocol involving age-appropriate sedentary, light and moderate-to vigorous-intensity physical activities. Children wore an ActiGraph GT3X on the right mid-axillary line of the hip. Energy expenditure measured by room calorimetry and physical activity intensity classified using direct observation were the criterion methods. Energy expenditure was predicted using Pate and Puyau equations. Physical activity intensity was classified using Evenson, Sirard, Van Cauwenberghe, Pate, Puyau, and Reilly, ActiGraph cut-points.

Results
The Pate equation significantly overestimated VO2 during sedentary behaviors, light physical activities and total VO2 (P<0.001). No difference was found between measured and predicted VO2 during moderate-to vigorous-intensity physical activities (P = 0.072). The Puyau equation significantly underestimated activity energy expenditure during moderate-to vigorous-intensity physical activities, light-intensity physical activities and total activity energy expenditure (P<0.0125). However, no overestimation of activity energy expenditure during sedentary behavior was found. The Evenson cut-point demonstrated significantly higher accuracy for classifying sedentary behaviors and light-intensity physical activities than others. Classification accuracy for moderate-to vigorous-intensity physical activities was significantly higher for Pate than others.

Conclusion
Available ActiGraph equations do not provide accurate estimates of energy expenditure across physical activity intensities in preschoolers. Cut-points of ≤25counts⋅15 s−1 and ≥420 counts⋅15 s−1 for classifying sedentary behaviors and moderate-to vigorous-intensity physical activities, respectively, are recommended.

Zeros of Dirichlet L-functions on the critical line with the accuracy of 40000 decimal places

The zeros of Dirichlet L-functions for various moduli and characters are being computed with very high accuracy on a cluster of workstations at Deakin University. This collection is growing to include more zeros (other moduli and characters).

Effects of type reduction algorithms on forecasting accuracy of IT2FLS models

Type reduction (TR) is one of the key components of interval type-2 fuzzy logic systems (IT2FLSs). Minimizing the computational requirements has been one of the key design criteria for developing TR algorithms. Often researchers give more rewards to computationally less expensive TR algorithms. This paper evaluates and compares five frequently used TR algorithms based on their contribution to the forecasting performance of IT2FLS models. Algorithms are judged based on the generalization power of IT2FLS models developed using them. Synthetic and real world case studies with different levels of uncertainty are considered to examine effects of TR algorithms on forecasts' accuracies. As per obtained results, Coupland-Jonh TR algorithm leads to models with a higher and more stable forecasting performance. However, there is no obvious and consistent relationship between the widths of the type reduced set and the TR algorithm. © 2013 Elsevier B.V.

The accuracy of Fastloc-GPS locations and implications for animal tracking

Over recent years, a major breakthrough in marine animal tracking has occurred with the advent of Fastloc-GPS that provides highly accurate location data even for animals that only surface briefly such as sea turtles, marine mammals and penguins. We assessed the accuracy of Fastloc-GPS locations using fixed trials of tags in which >45 000 locations were obtained. Procedures for determining the speed of travel and heading were developed by simulating tracks and then adding Fastloc-GPS location errors. The levels of detail achievable for speed and heading estimates were illustrated by using empirical Fastloc-GPS data for a green turtle (Chelonia mydas, Linnaeus, 1758) travelling over 3000 km across the Indian Ocean. The accuracy of Fastloc-GPS locations varied as a function of the number of GPS satellites used in the location calculation. For example, when Fastloc-GPS locations were calculated using 4 GPS satellites, 50% of locations were within 36 m and 95% within 724 m of the true position. These values improved to 18 and 70 m, respectively, when 6 satellites were used. Simulations indicated that for animals travelling around 2·5 km h-1 (e.g. turtles, penguins and seals) and depending on the number of satellites used in the location calculation, robust speed and heading estimates would usually be obtained for locations only 1-6 h apart. Fastloc-GPS accuracy is several orders of magnitude better that conventional Argos tracking or light-based geolocation and consequently will allow new insights into small-scale movement patterns of marine animals.

Validity of treadmill- and track-based individual calibration methods for estimating free-living walking speed and VO2 using the Actigraph accelerometer

BACKGROUND: For many patients clinical prescription of walking will be beneficial to health and accelerometers can be used to monitor their walking intensity, frequency and duration over many days. Walking intensity should include establishment of individual specific accelerometer count, walking speed and energy expenditure (VO2) relationships and this can be achieved using a walking protocol on a treadmill or overground. However, differences in gait mechanics during treadmill compared to overground walking may result in inaccurate estimations of free-living walking speed and VO2. The aims of this study were to compare the validity of track- and treadmill-based calibration methods for estimating free-living level walking speed and VO2 and to explain between-method differences in accuracy of estimation.

METHODS: Fifty healthy adults [32 women and 18 men; mean (SD): 40 (13) years] walked at four pre-determined speeds on an outdoor track and a treadmill, and completed three 1-km self-paced level walks while wearing an Actigraph monitor and a mobile oxygen analyser. Speed- and VO2-to-Actigraph count individual calibration equations were computed for each calibration method. Between-method differences in calibration equation parameters, prediction errors, and relationships of walking speed with VO2 and Actigraph counts were assessed. RESULTS: The treadmill-calibration equation overestimated free-living walking speed (on average, by 0.7 km · h(-1)) and VO2 (by 4.99 ml · kg(-1) · min(-1)), while the track-calibration equation did not. This was because treadmill walking, from which the calibration equation was derived, produced lower Actigraph counts and higher VO2 for a given walking speed compared to walking on a track. The prediction error associated with the use of the treadmill-calibration method increased with free-living walking speed. This issue was not observed when using the track-calibration method. CONCLUSIONS: The proposed track-based individual accelerometer calibration method can provide accurate and unbiased estimates of free-living walking speed and VO2 from walking. The treadmill-based calibration produces calibration equations that tend to substantially overestimate both VO2 and speed.