Search

A common practice to capture the non-stationary characteristics of the time series data in Artificial Neural Network (ANN) is by randomly dividing the whole set of available data into training, validation and testing, i.e. the data in validation and testing are represented in the training data. Consequently, the usability of the developed model on data not represented by the training data used during the network model development process is always doubtful. In this work, we present a back-propagation neural network (BNN) model trained using one-day history data to predict soil moisture data at 1 km resolution for two future dates. Specifically, high soil moisture values were observed in the training data while the testing data were characterized by drier conditions due to minimal or no rainfall. Our model uses separate mean and standard deviation statistics values from the training and testing data, respectively, to the z-normalized data. With data pre-processed using this method, the BNN model next uses a moving window of size 4 km × 4 km to capture the spatial variability of the soil moisture throughout the 40 km × 40 km study area. The coupling of the normalization and moving window method managed to achieve average soil moisture with Root Mean Square (RMSE) of 3.67% and correlation coefficient,R2of 0.89. By only using the suggested normalization without the moving window method, the BNN model managed to achieve an average RMSE of barely 5.82% withR2 = 0.83. When comparing with the normal practice of using the same mean and standard deviation statistics of the training data in the testing data, the retrieval accuracy of the BNN model deteriorates to 8.86% withR2 = 0.32. The experiment results demonstrated that the proposed coupling method performed better in terms of both RMSE and R2for soil moisture retrieval.

Traditionally the chemical industry uses coal, minerals and petroleum as its basic raw materials, but palm oil and palm kernel oil are being increasingly used as economical raw materials especially for the production of oleochemicals. High magnitude palm oil price volatility in recently years has been a major challenge faced the industry. Though many time series models have been developed, few have wide adoption in the industry, and one of the key issues is the sampling interval used in the models. To date, little effort has been spent on mining historical data to determine the representativeness of interval sampling. This paper presents a novel approach in identifying price equilibrium for crude palm oil by mining the sampling amount through historical price distribution. Evaluation is done on the outcomes of the experiment, and analysis is performed on the attributes of each different criteria of the price distribution. The performance of the proposed approach is also compared to the conventional Bollinger Band with static sampling size. Overall, the preliminary results show that price distribution with leptokurtic distribution outperforms other price distribution patterns, this will definitely assist further works to devise a novel financial time series analysis technique.

Spectrum-based Fault Localization (SBFL) has been widely studied as a debugging technique to reduce time and effort in locating faulty code in software. In SBFL, execution profiles (spectra) of pass and fail test cases are analyzed with SBFL metric to rank software code according to their likeliness to be faulty. However, there are significantly more pass test cases than fail test cases in typical test suites for faulty programs. The domination of pass test cases creates imbalance test suites that have a negative impact on the performance of SBFL. This is attributed to the fact that the execution profiles of fail test cases provide the essential information on the location of faulty code. In this paper, we propose to clone the spectra of fail test cases beyond balanced test suite to improve the performance of SBFL. Our empirical study shows that the proposed cloning method significantly improves the performance of commonly used SBFL metrics. Furthermore, we attempt to identify the amount of cloning required to achieve the optimal performance for the proposed cloning method.

Spectrum-based Fault Localization (SBFL) has been proven to be an effective technique to locate faulty statement in program code. SBFL metrics exploit the records of statement execution (spectra) by pass and fail test cases for each statement to rank its likeliness to be faulty. However, in some cases, these spectra contain duplicated and ambiguous information or noise which may deteriorate the performance of SBFL metrics. We propose six noise reduction schemes to eliminate test cases which provide duplicated and ambiguous information and evaluate the resulting performance improvements in SBFL metrics. Based on our findings, we further provide a guide for SBFL practitioners to select the best performing noise reduction scheme for the SBFL metrics that they use.

Spectrum-based Fault Localization (SBFL) is a well-known debugging technique that locates fault in program code by utilizing execution profiles (spectra) of pass and fail test cases. Hence, the performance of SBFL depends on the test cases executed and the test results. In the most extreme scenarios, the debugging process may have to be conducted with only one fail test case, one pass test case or no pass test case. These scenarios might occur due to extremely high or extremely low failure rates or when software testers decide to stop running more test cases due to time and resource constraints. However, limited test case execution profiles may reduce the accuracy of SBFL metrics. In view of this, we evaluate the performance of SBFL metrics in these extreme scenarios to identify the best performing SBFL metric for each of these scenarios. From the experiment results, we have further discovered the convergence in performance for SBFL metrics under these extreme scenarios.

Spectrum-based Fault Localization (SBFL) is a popular fault localization technique that ranks statements in a program according to their suspiciousness to be faulty based on the statement execution records (spectra) of pass and fail test cases. Many SBFL metrics have been proposed with varying accuracies in ranking of faulty statement. In this paper we proposed a new SBFL metric based on a pair scoring approach. We evaluated the performance of the proposed metric and compare it with other existing SBFL metrics. Despite its simplicity, we found the proposed metric outperformed majority of the existing SBFL metrics. 

Fault based testing aims at detecting hypothesized faults based on specifications or program source. There are some fault based techniques for testing Boolean expressions which are commonly used to model conditions in specifications as well as logical decisions in program source. The MUMCUT strategy has been proposed to generate test cases from Boolean expressions. Moreover, it detects eight common types of hypothesized faults provided that the original expression is in irredundant disjunctive normal form, IDNF. Software practitioners are more likely to write the conditions and logical decisions in general form rather than IDNF. Hence, it is interesting to investigate the fault detecting capability of the MUMCUT strategy with respect to general form Boolean expressions. In this article, we perform empirical studies to investigate the fault detection capability of the MUMCUT strategy with respect to general form Boolean expressions as well as mutated expressions. A mutated expression can be obtained from the original given Boolean expression by making a syntactic change based on a particular type of fault.

High magnitude of price fluctuation in crude palm oil has been a major challenge haunting small planters, plantation companies, investors, and all the up-stream and down-stream industry players. The severe price movements over the past few years has negatively impacted all the industry participants, especially in their decision making process concerning risk management and mitigation. This paper presents an approach of identifying price equilibrium for crude palm oil through the mining of historical price distribution, thus fostering the forecasting of potential high magnitude price movement upon the breaching of the price equilibrium. Experiment was conducted on historical price data of crude palm oil futures over the period of sixteen years to access its competency in identifying price equilibrium, and forecasting the magnitude of price movement after the breaching of price from the equilibrium level. Evaluation has been done to scrutinize the characteristic of price distribution, and the amount of price data used in forming the distribution. The outcome of the experiment reveals a promising performance demonstrating the capability to forecast high magnitude price movement of crude palm oil. This study constitutes a novel approach of using price distribution in price movement analysis, and this will aid the analysis process in financial decision making routine.

Financial charting software is widely used in share, commodity and foreign currency exchange markets to visualize and analyze price movements. Its quality is critical because incorrect outputs may lead to wrong analysis and trading decisions, and consequently substantial financial losses. Human visual judgment is often required to test financial charting software because of the graphical complexity of software outputs and limited knowledge of expected outputs. Such approach is labour intensive and error-prone. In this paper, we propose an automated testing technique combining metamorphic testing, assertion checking and a novel data label extraction method to eliminate human visual judgment from testing financial charting software. We used this technique to test pre-release builds of a commercial Point and Figure charting software component, and demonstrated that the proposed technique can effectively detect actual faults in the software component. Further, we discuss how the technique can be extended to test other charting software components.

Spectrum-based Fault Localization (SBFL) is an emerging debugging technique that assists software developers to locate faulty code in software. By utilizing code execution information (spectra), SBFL metrics rank lines of codes in software according to their likeliness to be faulty. However, recent studies showed that contradicting, duplicated or noisy spectra may deteriorate the ranking accuracy of SBFL metrics. In this paper, we propose and develop a novel SBFL tool with test case preprocessor to filter out test cases with contradicting, duplicated or other noisy spectra. Case studies conducted on real life faulty programs show that the proposed SBFL tool with test case preprocessor has successfully improved the performance of SBFL metrics in majority of the cases studied.

Fault-based testing has been proven to be a cost effective testing technique for software logics and rules expressed in Boolean expressions. It can guarantee the elimination of common faults without exhaustive testing. However, average software testing practitioners may not have in-depth knowledge on Boolean algebra and complex logic derivations required to apply existing fault-based testing techniques. In this paper, a dynamic fault visualization tool has been proposed. This tool allows its user to visualize fault-based testing and prioritize test inputs with a simple greedy method. The performance evaluation of this tool has been done on Boolean expressions extracted from a real life aviation tool. The results show that it can achieve significant performance improvements compared to ordinary sequential order test execution and existing static technique. The proposed visualization tool could also identify possible faults to guide the debugging process.

Software Testing is a costly and time consuming process in software development. Therefore, software testing tools are often deployed to automate the process in order to reduce cost and improve efficiency. However, many of them are proprietary and expensive. Hence, open source software testing tools could be an appealing alternative. In this paper, we survey the current states of open source software testing tools from three aspects, namely, their availability for different programming platforms and types testing activities, maintenance of the tools and license limitations. From the 152 tools surveyed, we found that open source software testing tools not only are widely available for popular programming platforms, but also support a wide range of testing activities. Furthermore, we also found that more than half of the tools surveyed have been actively maintained and updated by the open source communities. Finally, these tools have very few licensing limitations for commercial use, customization and redistribution.

Fuzzing is an automated black-box testing technique conducted with a destructive aim to crash (that is, to reveal failures in) the software under test. In this paper, we propose an adaptive random approach to fuzz the Out-Of-Memory (OOM) Killer on an embedded Linux distribution. The fuzzing process has revealed OOM Killer failures that cause the Linux kernel to remain in the OOM condition and become non-responsive. We have also found that the OOM Killer failures are more likely to occur when the Linux kernel has a higher over-commitment of memory requests. Finally, we have shown that the proposed adaptive random approach for fuzzing can reveal an OOM Killer failure with significantly fewer test inputs compared to the pure random approach.

Many financial trading and charting software packages provide users with technical indicators to analyze and predict price movements in financial markets. Any computation fault in technical indicator may lead to wrong trading decisions and cause substantial financial losses. Testing is a major software engineering activity to detect computation faults in software. However, there are two problems in testing technical indicators in these software packages. Firstly, the indicator values are updated with real-time market data that cannot be generated arbitrarily. Secondly, technical indicators are computed based on a large amount of market data. Thus, it is extremely difficult, if not impossible, to derive the expected indicator values to check the correctness of the computed indicator values. In this paper, we address the above problems by proposing a new testing technique to detect faults in computation of technical indicators. We show that the proposed technique is effective in detecting computation faults in faulty technical indicators on the MetaTrader 4 Client Terminal.

Usual techniques used in the automatic test case generation and executions assume that a complete oracle will be available during the testing process. However, in some numerical programs this assumption is not valid because the oracle in form of analytical solution is either not available or too difficult to obtain. To overcome this problem, metamorphic testing proposes to generate follow-up test cases to check important properties of the output function. We study the intrinsic properties of Medial Axes for casting applications. Even though Medial Axes is a geometric function that is independent from the output function of casting simulation, we propose to use its intrinsic properties for metamorphic testing of casting simulation to alleviate the oracle problem.