.......................................................................................................5 Introduction ..................................................................................................6 Methods........................................................................................................7 Results ........................................................................................................10 Discussion ..................................................................................................11 Conclusion .................................................................................................12 Acknowledgements ....................................................................................12 References ..................................................................................................13 3 AUTOMATING THE DETECTION OF NEGATIVE APPENDICITIS IN PATHOLOGY REPORTS USING NATURAL LANGUAGE PROCESSING ................................................................................................21 Abstract ......................................................................................................22.....................................................................................................22

Computing Education has become a widely popular research field at top universities with much attention aimed at improving and expanding K-12 computer science education. Though numerous efforts are being made by institutions, industries, and the community, many challenges still prevent widespread K-12 CS education. Our research aims to alleviate these challenges with a new adaptive learning system to teach introductory programming in a unique and interesting way. Adaptive learning strategies normally adapt based on a student's previous knowledge, pace, or learning style. Our research takes a new approach to adapt the content, practice problems, and examples based on a student's interests. Interest-based learning has been shown to improve intrinsic motivation, leading to better learning and achievements. This paper outlines how SAIL - a System for Adaptive Interest-based Learning - could impact introductory CS education and alleviate many of its challenges.

A new search engine, based on Hooke's Law, is proposed and applied to both the Multiple Subscriber Equipment (MSE) deployment and Schaffer's F6 equation problems. A virtualized Damped Simple Harmonic Oscillator is constructed using a system with multiple oscillating. The VHO search engine is constructed to converge on an optimal or near-optimal solution, and then, due to damping, come to rest, at which point the search stops. For the MSE problem, the engine finds the optimal solution 90 to nearly 100% of the time (depending on settings) and does so efficiently. For Schaffer's F6, the engine produces more variable fitness levels than a reference Genetic Algorithm (GA) search engine, but has found two solutions better than the GA.

The objective of the current study was to develop a solar radiation forecasting model capable of determining the specific times during a given day that solar panels could be relied upon to produce energy in sufficient quantities to meet the demand of the energy provider, Southern Company. Model averaged neural networks (MANN) and alternating model trees (AMT) were constructed to forecast solar radiation an hour into the future, given 2003–2012 solar radiation data from the Griffin, GA weather station for training and 2013 data for testing. Generalized linear models (GLM), random forests, and multilayer perceptron (MLP) were developed, in order to assess the relative performance improvement attained by the MANN and AMT models. In addition, a literature review of the most prominent hourly solar radiation models was performed and normalized root mean square error was calculated for each, for comparison with the MANN and AMT models. The results demonstrate that MANN and AMT models outperform or parallel the highest performing forecasting models within the literature. MANN and AMT are thus promising time series forecasting models that may be further improved by combining these models into an ensemble.

The optimization of a water distribution network (WDN) is a highly nonlinear, multi-modal, and constrained combinatorial problem. Particle swarm optimization (PSO) has been shown to be a fast converging algorithm for WDN optimization. An improved estimation of distribution algorithm (EDA) using historic best positions to construct a sample space is hybridized with PSO both in sequential and in parallel to improve population diversity control and avoid premature convergence. Two water distribution network benchmark examples from the literature are adopted to evaluate the performance of the proposed hybrid algorithms. The experimental results indicate that the proposed algorithms achieved the literature record minimum (6.081 M$) for the small size Hanoi network. For the large size Balerma network, the parallel hybrid achieved a slightly lower minimum (1.921M€) than the current literature reported best minimum (1.923M€). The average number of evaluations needed to achieve the minimum is one order smaller than most existing algorithms. With a fixed, small number of evaluations, the sequential hybrid outperforms the parallel hybrid showing its capability for fast convergence. The fitness and diversity of the populations were tracked for the proposed algorithms. The track record suggests that constructing an EDA sample space with historic best positions can improve diversity control significantly. Parallel hybridization also helps to improve diversity control yet its effect is relatively less significant.

Computer Science (CS) seems to be everywhere - in our smartphones, apps, cars, watches and so much more. It is integrated into almost every discipline and has a growing importance in our daily lives, yet even with top salaries, exciting companies, and promising careers, the job market remains underpopulated and under-representative of women. Though its applications are everywhere, CS is an often misunderstood field due to lack of exposure and knowledge. This paper explores the misperception associated with Computer Science and examines the efforts to make it more inclusive by attracting women, introducing CS earlier at the K-12 level, and improving entry-level college courses.

The Jamaica Stock Exchange (JSE) has been defined by Standard and Poor’s as a frontier market. It has undergone periods where trading gains exceeded that of major markets such as the London Stock Exchange. The randomness of the JSE was investigated over the period 2001–2014, using statistical tests and the Hurst exponent to reveal periods when the JSE did not follow a random walk. This chapter focuses on machine learning algorithms including decision trees, neural networks and support vector machines used to predict the JSE. Selected algorithms were applied to trading data over a 22 month period for price and trend forecasting and a 12-year period for volume forecasts. Experimental results show 90 % accuracy in the movement prediction with mean absolute error of 0.4 and 0.95 correlation coefficient for price prediction. Volume predictions were enhanced by a discretization method and support vector machine to yield over 70 % accuracy.

A new search engine, based on Hooke's Law, is proposed and applied to both the Multiple Subscriber Equipment (MSE) deployment and Schaffer's F6 equation problems. A virtualized Damped Simple Harmonic Oscillator is constructed using a system with multiple oscillating. The VHO search engine is constructed to converge on an optimal or near-optimal solution, and then, due to damping, come to rest, at which point the search stops. For the MSE problem, the engine finds the optimal solution 90 to nearly 100% of the time (depending on settings) and does so efficiently. For Schaffer's F6, the engine produces more variable fitness levels than a reference Genetic Algorithm (GA) search engine, but has found two solutions better than the GA.

The Jamaica Stock Exchange (JSE) has been defined by Standard and Poor's as a frontier market. It has undergone periods where trading gains exceeded that of major markets such as the London Stock Exchange. This paper focuses on machine learning algorithms including decision trees, neural networks and support vector machines used to predict the JSE. Selected algorithms were applied to trading data over a 22 month period for price and trend forecasting and a 12-year period for volume forecasts. Experimental results show 90% accuracy in the movement prediction and 0.95 correlation coefficient for price prediction. Volume predictions were enhanced by a discretization method and support vector machine to yield over 70% accuracy.

Since the problem's formulation by Kautz in 1958 as an error detection tool, diverse applications for long snakes and coils have been found. These include coding theory, electrical engineering, and genetics. Over the years, the problem has been explored by many researchers in different fields using varied approaches, and has taken on additional meaning. The problem has become a benchmark for evaluating search techniques in combinatorially expansive search spaces (NP-complete Optimizations).We build on our previous work and present improved heuristics for Stochastic Beam Search sub-solution selection in searching for longest induced paths: open (snakes), closed (coils), and symmetric closed (symmetric coils); in n-dimensional hypercube graphs. Stochastic Beam Search, a non-deterministic variant of Beam Search, provides the overall structure for our search. We present eleven new lower bounds for the Snake-in-the-Box problem for snakes in dimensions 11, 12, and 13; coils in dimensions 10, 11, and 12; and symmetric coils in dimensions 9, 10, 11, 12, and 13. The best known solutions of the unsolved dimensions of this problem have improved over the years and we are proud to make a contribution to this problem as well as the continued progress in combinatorial search techniques.

The Snake in the Box problem is an NP-Hard problem. The goal is to find the longest maximal snakes (a certain kind of path satisfying particular constraints described as “spread”) in an n-dimensional hypercube [8]. With increasing dimensions the search space grows exponentially and the search for snakes becomes more and more difficult. This article identifies an underlying pattern among the known longest snakes in previously searched dimensions, which resembles the DNA of living cells in many ways. Surprisingly, these generic structures are fundamentally different for the four combinations of odd and even dimension and spread. It briefly explains the reason why they have different underlying structures. In odd dimensions with odd spread, there is one symmetric point and a unique mapping of complementary transition pairs and are discussed in detail in this paper. This article focusses only on one of these – odd dimension with odd spread. Later, it also reports three new lower bounds that are established using these generic structures from previously known longest maximal snakes. Another known longest snake in another odd dimension with odd spread is also found using this approach.

Social media comments have in the past had an instantaneous effect on stock markets. This paper investigates the sentiments expressed on the social media platform Twitter and their predictive impact on the Jamaica Stock Exchange. A hybrid predictive model of sentiment analysis and machine learning algorithms including decision trees, neural networks and support vector machines are used to predict the Jamaica Stock Exchange. The architecture created, SentAMaL, investigated the impact of sentiments on medical marijuana legalization on relevant stock indices. Due to the unstructured nature of tweets, a customized preprocessing routine was developed prior to determining sentiment and to perform the prediction. Experimental results show 87% accuracy in the movement prediction and 0.99 correlation coefficient for price prediction.

This research represents a novel application of operations research methodology to a form of a hierarchical open vehicle routing problem. In our case study, a road network was divided into priority sets such that each set needed to be completely addressed before work could begin on the next. Rule-based heuristics and adaptations of local beam search heuristics were tested in various combinations, viable solutions were developed, and results were evaluated based on the time required to clear the entire road network of storm-generated debris. The results were also compared against a theoretical lower bound on the time required to clear all roads. The best routes were generated using a combination of constant time beam search and a rule-based heuristic, 12.4% greater than the lower bound, yet understandable for the case study area that contained numerous dead-ends and disconnected road priority sets.

Since the problem's formulation by Kautz in 1958 as an error detection tool, diverse applications for long snakes and coils have been found. These include coding theory, electrical engineering, and genetics. Over the years, the problem has been explored by many researchers in different fields using varied approaches, and has taken on additional meaning. The problem has become a benchmark for evaluating search techniques in combinatorially expansive search spaces (NP-complete Optimizations).We present an effective process for searching for long achordal open paths (snakes) and achordal closed paths (coils) in n-dimensional hypercube graphs. Stochastic Beam Search provides the overall structure for the search while graph theory based techniques are used in the computation of a generational fitness value. This novel fitness value is used in guiding the search. We show that our approach is likely to work in all dimensions of the SIB problem and we present new lower bounds for a snake in dimension 11 and coils in dimensions 10, 11, and 12. The best known solutions of the unsolved dimensions of this problem have improved over the years and we are proud to make a contribution to this problem as well as the continued progress in combinatorial search techniques.

Predicted air and dew point temperatures can be valuable in decision making in many areas including protecting crops from damage, avoiding heat stress on animals and humans, and in planning related to energy management. Current web-based artificial neural network (ANN) models on the Automated Environment Monitoring Network (AEMN) in Georgia predict hourly air and dew point temperature for twelve prediction horizons, using 24 models. The observed air temperature may approach the observed dew point temperature, but never goes below it. Current web based ANN models have prediction errors which, when the air and dew point temperatures are close, may cause air temperature to be predicted below the dew point temperature. Herein this error is referred to as a prediction anomaly. The goal of this research was to improve the prediction accuracy of existing air and dew point temperature ANN models by combining the two weather variables into a single ANN model for each prediction horizon. The objectives of this study were to reduce the mean absolute error (MAE) of prediction and to reduce the number of prediction anomalies. The combined models produced a reduction in the air temperature MAE for ten of twelve prediction horizons with an average reduction in MAE of 1.93 %. The combined models produced a reduction in the dew point temperature MAE for only six of twelve prediction horizons with essentially no average decrease in MAE. However, the combined models showed a marked reduction in prediction anomalies for all twelve prediction horizons with an average reduction of 34.1 %. The reduction in prediction anomalies ranged from 4.6 % at the one-hour horizon to 60.5 % at the eleven-hour horizon.

Parameter setting is very essential for the application of particle swarm optimization (PSO), especially the acceleration coefficients. In this paper, we propose a fast estimation strategy of optimal parameter setting for PSO, in which an estimation distribution algorithm (EDA) is used to co-evolve the acceleration coefficients (!!and !!). The proposed algorithm is validated on two numerical optimization problems and then applied to the urban water distribution network optimization problem. The experimental results show that both of these two parameters converge to a fixed value respectively and the achieved values for !! and !! are consistent as the results of parameter tuning.PSO with the estimated optimal parameters could achieve the best solution on benchmark example and also outperform other methods in terms of reliability and efficiency.

The accurate prediction of air temperature is important in many areas of decision-making including agricultural management, transportation and energy management. Previous research has focused on the development of artificial neural network (ANN) models to predict air temperature from one to twelve hours in advance. The inputs to these models included a constant duration of prior data with a fixed resolution for all environmental variables for all prediction horizons. The overall goal of this research was to develop more accurate ANN models that could predict air temperature for each prediction horizon. The specific objective was to determine if the ANN model accuracy could be improved by applying a genetic algorithm (GA) for each prediction horizon to determine the preferred duration and resolution of input prior data for each environmental variable. The ANN models created based on this GA based approach provided smaller errors than the models created based on the existing constant duration and fixed data resolution approach for all twelve prediction horizons. Except for a few cases, the GA generally included a longer duration for prior air temperature data and shorter durations for other environmental variables. The mean absolute errors (MAEs) for the evaluation input patterns of the one-, four-, eight-, and twelve-hour prediction models that were based on this GA approach were 0.564°C, 1.264°C, 1.766°C and 2.018°C, respectively. These MAEs were improvements of 3.98%, 4.59%, 2.55% and 1.70% compared to the models that were created based on the existing approach for the same corresponding prediction horizons. Thus, the GA based approach to determine the duration and resolution of prior input data resulted in more accurate ANN models than the existing ones for air temperature prediction. Future work could examine the effects of various GA and fitness evaluation parameters that were part of the approach used in this study.