Prediction of wool knitwear pilling propensity using data mining techniques

This thesis examined the application of data mining techniques to the issue of predicting pilling propensity of wool knitwear. Using real industrial data, a pilling propensity prediction tool with embedded trained support vector machines is developed to provide high accuracy prediction to wool knitwear even before the yarn is spun!

Data mining techniques for the assessment of factors contributing to the damage of residential houses in Australia

This paper reports on the preparation and management processes of inconsistent data on damage on residential houses in Victoria, Australia. There are no existing specific and fully relevant databases readily available except for the incomplete paper-based and electronic-based reports. Therefore, the extracting of information from the reports is complicated and time consuming in order to extract and include all the necessary information needed for analysis of damage on residential houses founded on expansive soils. Data mining is adopted to develop a database. Statistical methods and Artificial Intelligence methods are used to quantify the quality of data. The paper concludes that the development of such database could enable BHC to evaluate the usefulness of the reports prepared on the reported damage properties for further analysis.

An examination of Australian gold mining firms' exposure over the collapse of gold price in the late 1990s

Purpose – The purpose of this study is to examine the exposures of Australian gold mining firms in the highly volatile period from 1995 to 2000. This period has been characterized by significant changes in gold price due to bulk sale of gold by collective central banks. Specifically, the paper aims to investigate several firm-specific factors that are hypothesized to carry substantial influence on gold beta.

Design/methodology/approach – To estimate gold beta, we use the following multifactor model: R_g,_t = a+ß_gGPR_t + ß_xFXR_t + ß_mR_m,_t + _Et , where R_g,_t is the return on the gold stock Index at time t, GPR_t is the gold price return denominated in US dollar at time t, FXR_t is the foreign exchange return of Australian dollar in terms of US dollar at time t, R_m,t is the market return at time t, and _Et is the random error term at time t.

Findings – The paper finds that the values of gold beta are consistently greater than one, implying the sensitive nature of firms’ stock returns to gold price changes. This also suggests that investors holding gold mining stock would receive higher percentage increases in stock returns from a percentage increase in gold price returns, as opposed to investors holding gold bullion. Furthermore, these values have changed substantially over time with significant changes in gold price volatility. The most important and consistent relationship that we find is the impact of firms’ hedging behavior on their respective gold betas. This is consistent with Tufano’s study. It implies that firms, which hedge a greater proportion of their gold reserves, are less sensitive to movements in gold prices. The finding therefore supports the risk management theory that hedging increases shareholder’s wealth. However, cash operating costs, cash reserves and the level of gold production seem to influence very little on the firms’ exposure to gold price changes.

Originality/value – This study is of interest and important to the stock mining companies and investors because the extent of the effect of gold price movements on the stock returns of gold mining companies has significant impacts on returns for both firms and investors especially in their risk management and investment decisions, respectively.

Sprint cycling performance is maintained with short-term contrast water immersion

Purpose: Given the widespread use of water immersion during recovery from exercise, we aimed to investigate the effect of contrast water immersion on recovery of sprint cycling performance, HR and, blood lactate.

Methods: Two groups completed high-intensity sprint exercise before and after a 30-min randomized recovery. The Wingate group (n = 8) performed 3 x 30-s Wingate tests (4-min rest periods). The repeated intermittent sprint group (n = 8) cycled for alternating 30-s periods at 40% of predetermined maximum power and 120% maximum power, until exhaustion. Both groups completed three trials using a different recovery treatment for each trial (balanced randomized application). Recovery treatments were passive rest, 1:1 contrast water immersion (2.5 min of cold (8-C) to 2.5 min of hot (40-C)), and 1:4 contrast water immersion (1 min of cold to 4 min of hot). Blood lactate and HR were recorded throughout, and peak power and total work for pre- and postrecovery Wingate performance and exercise time and total work for repeated sprinting were recorded.

Results: Recovery of Wingate peak power was 8% greater after 1:4 contrast water immersion than after passive rest, whereas both contrast water immersion ratios provided a greater recovery of exercise time (È10%) and total work (È14%) for repeated sprinting than for passive rest. Blood lactate was similar between trials. Compared with passive rest, HR initially declined more slowly during contrast water immersion but increased with each transition to a cold immersion phase.

Conclusions: These data support contrast water immersion being effective in maintaining performance during a short-term recovery from sprint exercise. This effect needs further investigation but is likely explained by cardiovascular mechanisms, shown here by an elevation in HR upon each cold immersion.

Fast mining of non-derivable episode rules in complex sequences

Researchers have been endeavoring to discover concise sets of episode rules instead of complete sets in sequences. Existing approaches, however, are not able to process complex sequences and can not guarantee the accuracy of resulting sets due to the violation of anti-monotonicity of the frequency metric. In some real applications, episode rules need to be extracted from complex sequences in which multiple items may appear in a time slot. This paper investigates the discovery of concise episode rules in complex sequences. We define a concise representation called non-derivable episode rules and formularize the mining problem. Adopting a novel anti-monotonic frequency metric, we then develop a fast approach to discover non-derivable episode rules in complex sequences. Experimental results demonstrate that the utility of the proposed approach substantially reduces the number of rules and achieves fast processing.

Improving reliability of unbalanced text mining by reducing performance bias

Class imbalance in textual data is one important factor that affects the reliability of text mining. For imbalanced textual data, conventional classifiers tend to have a strong performance bias, which results in high accuracy rate on the majority class but very low rate on the minorities. An extreme strategy for unbalanced learning is to discard the majority instances and apply one-class classification to the minority class. However, this could easily cause another type of bias, which increases the accuracy rate on minorities by sacrificing the majorities. This chapter aims to investigate approaches that reduce these two types of performance bias and improve the reliability of discovered classification rules. Experimental results show that the inexact field learning method and parameter optimized one class classifiers achieve more balanced performance than the standard approaches.

Maximum contrast analysis for nonnegative blind source separation

In this paper, we propose a maximum contrast analysis (MCA) method for nonnegative blind source separation, where both the mixing matrix and the source signals are nonnegative. We first show that the contrast degree of the source signals is greater than that of the mixed signals. Motivated by this observation, we propose an MCA-based cost function. It is further shown that the separation matrix can be obtained by maximizing the proposed cost function. Then we derive an iterative determinant maximization algorithm for estimating the separation matrix. In the case of two sources, a closed-form solution exists and is derived. Unlike most existing blind source separation methods, the proposed MCA method needs neither the independence assumption, nor the sparseness requirement of the sources. The effectiveness of the new method is illustrated by experiments using X-ray images, remote sensing images, infrared spectral images, and real-world fluorescence microscopy images.

Semantic trajectory based event detection and event pattern mining

Video event detection is an effective way to automatically understand the semantic content of the video. However, due to the mismatch between low-level visual features and high-level semantics, the research of video event detection encounters a number of challenges, such as how to extract the suitable information from video, how to represent the event, how to build up reasoning mechanism to infer the event according to video information. In this paper, we propose a novel event detection method. The method detects the video event based on the semantic trajectory, which is a high-level semantic description of the moving object’s trajectory in the video. The proposed method consists of three phases to transform low-level visual features to middle-level raw trajectory information and then to high-level semantic trajectory information. Event reasoning is then carried out with the assistance of semantic trajectory information and background knowledge. Additionally, to release the users’ burden in manual event definition, a method is further proposed to automatically discover the event-related semantic trajectory pattern from the sample semantic trajectories. Furthermore, in order to effectively use the discovered semantic trajectory patterns, the associative classification-based event detection framework is adopted to discover the possibly occurred event. Empirical studies show our methods can effectively and efficiently detect video events.

Infrequent item mining in multiple data streams

The problem of extracting infrequent patterns from streams and building associations between these patterns is becoming increasingly relevant today as many events of interest such as attacks in network data or unusual stories in news data occur rarely. The complexity of the problem is compounded when a system is required to deal with data from multiple streams. To address these problems, we present a framework that combines the time based association mining with a pyramidal structure that allows a rolling analysis of the stream and maintains a synopsis of the data without requiring increasing memory resources. We apply the algorithms and show the usefulness of the techniques. © 2007 Crown Copyright.