This paper considers four approaches to ordinal predictors in linear regression to evaluate how these contrast with respect to predictive accuracy. The two most typical treatments, namely, dummy coding and classic linear regression on assigned level scores are compared with two improved methods; penalized smoothed coefficients and a generalized additive model with cubic splines. A simulation study is conducted to assess all on the basis of predictive performance. Our results show that the dummy based methods surpass the numeric at low sample sizes. Although, as sample size increases, differences between the methods diminish. Tendencies of overfitting are identified among the dummy methods. We conclude by stating that the choice of method not only ought to be context driven, but done in the light of all characteristics.

Context. Reports suggest that defects in code cost the US in excess of $50billion per year to put right. Defect Prediction is an important part of Software Engineering. It allows developers to prioritise the code that needs to be inspected when trying to reduce the number of defects in code. A small change in the number of defects found will have a significant impact on the cost of producing software. Aims. The aim of this dissertation is to investigate the factors which a ect the performance of defect prediction models. Identifying the causes of variation in the way that variables are computed should help to improve the precision of defect prediction models and hence improve the cost e ectiveness of defect prediction. Methods. This dissertation is by published work. The first three papers examine variation in the independent variables (code metrics) and the dependent variable (number/location of defects). The fourth and fifth papers investigate the e ect that di erent learners and datasets have on the predictive performance of defect prediction models. The final paper investigates the reported use of di erent machine learning approaches in studies published between 2000 and 2010. Results. The first and second papers show that independent variables are sensitive to the measurement protocol used, this suggests that the way data is collected a ects the performance of defect prediction. The third paper shows that dependent variable data may be untrustworthy as there is no reliable method for labelling a unit of code as defective or not. The fourth and fifth papers show that the dataset and learner used when producing defect prediction models have an e ect on the performance of the models. The final paper shows that the approaches used by researchers to build defect prediction models is variable, with good practices being ignored in many papers. Conclusions. The measurement protocols for independent and dependent variables used for defect prediction need to be clearly described so that results can be compared like with like. It is possible that the predictive results of one research group have a higher performance value than another research group because of the way that they calculated the metrics rather than the method of building the model used to predict the defect prone modules. The machine learning approaches used by researchers need to be clearly reported in order to be able to improve the quality of defect prediction studies and allow a larger corpus of reliable results to be gathered.

The use of remote sensing technology has made it possible for the non-contact measurement of soil moisture content (SMC). Many remote sensing techniques can be used such as microwave sensors, electromagnetic waves sensors, capacitance, and thermal infrared sensors. Some of those techniques are constrained by their high fabrication cost, operation cost, size, or complexity. In this study, a thermal infrared technique was used to predict soil moisture content with the aid of using weather meteorological variables. <p>The measured variables in the experiment were soil moisture content (%SMC), soil surface temperature (Ts) measured using thermocouples, air temperature (Ta), relative humidity (RH), solar radiation (SR), and wind speed (WS). The experiment was carried out for a total of 12 soil samples of two soil types (clay/sand) and two compaction levels (compacted/non-compacted). After data analysis, calibration models relating soil moisture content (SMC) to differential temperature (Td), relative humidity (RH), solar radiation (SR), and wind speed (WS) were generated using stepwise multiple linear regression of the calibration data set. The performance of the models was evaluated using validation data. Four mathematical models of predicting soil moisture content were generated for each soil type and configuration using the calibration data set. Among the four models, the best model for each soil type and configuration was determined by comparing root mean of squared errors of calibration (RMSEC) and root mean of squared errors of validation (RMSEV) values. Furthermore, a calibration model for the thermal infrared sensor was developed to determine the corrected soil surface temperature as measured by the sensor (Tir) instead of using the thermocouples. The performance of the thermal infrared sensor to predict soil moisture content was then tested for sand compacted and sand non-compacted soils and compared to the predictive performance of the thermocouples. This was achieved by using the measured soil surface temperature by the sensor (Tir), instead of the measured soil surface temperature using the thermocouples to determine the soil-minus-air temperature (Td). The sensor showed comparable prediction performance, relative to thermocouples. <p>Overall, the models developed in this study showed high prediction performance when tested with the validation data set. The best models to predict SMC for compacted clay soil, non-compacted clay soil, and compacted sandy soil were three-variable models containing three predictive variables; Td, RH, and SR. On the other hand, the best model to predict SMC for compacted sandy soil was a two-variable model containing Td, and RH. The results showed that the prediction performance of models for predicting SMC for the sandy soils was superior to those of clay soils.

The use of remote sensing technology has made it possible for the non-contact measurement of soil moisture content (SMC). Many remote sensing techniques can be used such as microwave sensors, electromagnetic waves sensors, capacitance, and thermal infrared sensors. Some of those techniques are constrained by their high fabrication cost, operation cost, size, or complexity. In this study, a thermal infrared technique was used to predict soil moisture content with the aid of using weather meteorological variables. <p>The measured variables in the experiment were soil moisture content (%SMC), soil surface temperature (Ts) measured using thermocouples, air temperature (Ta), relative humidity (RH), solar radiation (SR), and wind speed (WS). The experiment was carried out for a total of 12 soil samples of two soil types (clay/sand) and two compaction levels (compacted/non-compacted). After data analysis, calibration models relating soil moisture content (SMC) to differential temperature (Td), relative humidity (RH), solar radiation (SR), and wind speed (WS) were generated using stepwise multiple linear regression of the calibration data set. The performance of the models was evaluated using validation data. Four mathematical models of predicting soil moisture content were generated for each soil type and configuration using the calibration data set. Among the four models, the best model for each soil type and configuration was determined by comparing root mean of squared errors of calibration (RMSEC) and root mean of squared errors of validation (RMSEV) values. Furthermore, a calibration model for the thermal infrared sensor was developed to determine the corrected soil surface temperature as measured by the sensor (Tir) instead of using the thermocouples. The performance of the thermal infrared sensor to predict soil moisture content was then tested for sand compacted and sand non-compacted soils and compared to the predictive performance of the thermocouples. This was achieved by using the measured soil surface temperature by the sensor (Tir), instead of the measured soil surface temperature using the thermocouples to determine the soil-minus-air temperature (Td). The sensor showed comparable prediction performance, relative to thermocouples. <p>Overall, the models developed in this study showed high prediction performance when tested with the validation data set. The best models to predict SMC for compacted clay soil, non-compacted clay soil, and compacted sandy soil were three-variable models containing three predictive variables; Td, RH, and SR. On the other hand, the best model to predict SMC for compacted sandy soil was a two-variable model containing Td, and RH. The results showed that the prediction performance of models for predicting SMC for the sandy soils was superior to those of clay soils.

This paper proposes an innovative way to address real cases of production prediction. This approach consists in the decomposition of original time series into time sub-series according to a group of factors in order to generate a predictive model from the partial predictive models of the sub-series. The adjustment of the models is carried out by means of a set of statistic techniques and Automatic Learning. This method was compared to an intuitive method consisting of a direct prediction of time series. The results show that this approach achieves better predictive performance than the direct way, so applying a decomposition method is more appropriate for this problem than non-decomposition.

No description available.

Processor architectures are becoming increasingly complex and hence architects have to evaluate a large design space consisting of several parameters, each with a number of potential settings. In order to assist in guiding design decisions we develop simple and accurate models of the superscalar processor design space using a detailed and validated superscalar processor simulator. Firstly, we obtain precise estimates of all significant micro-architectural parameters and their interactions by building linear regression models using simulation based experiments. We obtain good approximate models at low simulation costs using an iterative process in which Akaike’s Information Criteria is used to extract a good linear model from a small set of simulations, and limited further simulation is guided by the model using D-optimal experimental designs. The iterative process is repeated until desired error bounds are achieved. We use this procedure for model construction and show that it provides a cost effective scheme to experiment with all relevant parameters. We also obtain accurate predictors of the processors performance response across the entire design-space, by constructing radial basis function networks from sampled simulation experiments. We construct these models, by simulating at limited design points selected by latin hypercube sampling, and then deriving the radial neural networks from the results. We show that these predictors provide accurate approximations to the simulator’s performance response, and hence provide a cheap alternative to simulation while searching for optimal processor design points.