A Taxonomy of Big Data for Optimal Predictive Machine Learning and Data Mining

Big data comes in various ways, types, shapes, forms and sizes. Indeed, almost all areas of science, technology, medicine, public health, economics, business, linguistics and social science are bombarded by ever increasing flows of data begging to be analyzed efficiently and effectively. In this paper, we propose a rough idea of a possible taxonomy of big data, along with some of the most commonly used tools for handling each particular category of bigness. The dimensionality p of the input space and the sample size n are usually the main ingredients in the characterization of data bigness. The specific statistical machine learning technique used to handle a particular big data set will depend on which category it falls in within the bigness taxonomy. Large p small n data sets for instance require a different set of tools from the large n small p variety. Among other tools, we discuss Preprocessing, Standardization, Imputation, Projection, Regularization, Penalization, Compression, Reduction, Selection, Kernelization, Hybridization, Parallelization, Aggregation, Randomization, Replication, Sequentialization. Indeed, it is important to emphasize right away that the so-called no free lunch theorem applies here, in the sense that there is no universally superior method that outperforms all other methods on all categories of bigness. It is also important to stress the fact that simplicity in the sense of Ockham’s razor non-plurality principle of parsimony tends to reign supreme when it comes to massive data. We conclude with a comparison of the predictive performance of some of the most commonly used methods on a few data sets.

A Comparative Analysis of Predictive Learning Algorithms on High-Dimensional Microarray Cancer Data

This research evaluates pattern recognition techniques on a subclass of big data where the dimensionality of the input space (p) is much larger than the number of observations (n). Specifically, we evaluate massive gene expression microarray cancer data where the ratio κ is less than one. We explore the statistical and computational challenges inherent in these high dimensional low sample size (HDLSS) problems and present statistical machine learning methods used to tackle and circumvent these difficulties. Regularization and kernel algorithms were explored in this research using seven datasets where κ < 1. These techniques require special attention to tuning necessitating several extensions of cross-validation to be investigated to support better predictive performance. While no single algorithm was universally the best predictor, the regularization technique produced lower test errors in five of the seven datasets studied.

Collaborative design of assessment criteria to improve undergraduate student engagement and performance

Student engagement is vital in enhancing the student experience and encouraging deeper learning. Involving students in the design of assessment criteria is one way in which to increase student engagement. In 2011, a marking matrix was used at Aston University (UK) for logbook assessment (Group One) in a project-based learning module. The next cohort of students in 2012 (Group Two) were asked to collaboratively redesign the matrix and were given a questionnaire about the exercise. Group Two initially scored a lower average logbook mark than Group One. However, Group Two showed the greatest improvement between assessments, and the quality of, and commitment to, logbooks was noticeably improved. Student input resulted in a more defined, tougher mark scheme. However, this provided an improved feedback system that gave more scope for self-improvement. The majority of students found the exercise incorporated their ideas, enhanced their understanding, and was useful in itself.

Using observation ageing to improve Markovian model learning in QoS engineering

Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

Human resource management, innovation and performance

Human Resource Management, Innovation and Performance investigates the relationship between HRM, innovation and performance. Taking a multi-level perspective the book reflects critically on contentious themes such as high performance work systems, organizational design options, cross-boundary working, leadership styles and learning at work.

Learning user and product distributed representations using a sequence model for sentiment analysis

In product reviews, it is observed that the distribution of polarity ratings over reviews written by different users or evaluated based on different products are often skewed in the real world. As such, incorporating user and product information would be helpful for the task of sentiment classification of reviews. However, existing approaches ignored the temporal nature of reviews posted by the same user or evaluated on the same product. We argue that the temporal relations of reviews might be potentially useful for learning user and product embedding and thus propose employing a sequence model to embed these temporal relations into user and product representations so as to improve the performance of document-level sentiment analysis. Specifically, we first learn a distributed representation of each review by a one-dimensional convolutional neural network. Then, taking these representations as pretrained vectors, we use a recurrent neural network with gated recurrent units to learn distributed representations of users and products. Finally, we feed the user, product and review representations into a machine learning classifier for sentiment classification. Our approach has been evaluated on three large-scale review datasets from the IMDB and Yelp. Experimental results show that: (1) sequence modeling for the purposes of distributed user and product representation learning can improve the performance of document-level sentiment classification; (2) the proposed approach achieves state-of-The-Art results on these benchmark datasets.

Systemic country differences in the European innovation performance - Does institutional context matter?

The so-called "High Performance Working System" (HPWS) and the lean production are representing the theoretical and methodological foundations of this paper. In this relation it is worth making distinction between various theoretical streams of the HPWS. The first theoretical stream in the literature is focusing on the diffusion of the Japanese-style management and organizational practices both in the US and in the Europe. The second theoretical strand comprises the approach of sociology of work and dealing with the learning/innovation capabilities of the new forms of work organization. Finally, the third theoretical approach is addressing on the types of knowledge and learning process and their relations with the innovation capabilities of the firm. The authors’ analysis is based on the international comparison, both in regional and in cross country comparison. For regional comparison the share of ICT clusters in Europe, USA and the rest of the world was assessed. For the purpose of the cross-country comparison in the EU, the innovation performance measured by the index Innovation Union Scoreboard (IUS) was used in both the before and after the financial crisis.

Mindmapping as an adult learning notetaking strategy to increase the encoding of information in a corporate training course

A study was conducted to investigate the effectiveness, as measured by performance on course posttests, of mindmapping versus traditional notetaking in a corporate training class. The purpose of this study was to increase knowledge concerning the effectiveness of mindmapping as an information encoding tool to enhance the effectiveness of learning. Corporations invest billions of dollars, annually, in training programs. Given this increased demand for effective and efficient workplace learning, continual reliance on traditional notetaking is questionable for the high-speed and continual learning required on workers.^ An experimental, posttest-only control group design was used to test the following hypotheses: (1) there is no significant difference in posttest scores on an achievement test, administered immediately after the course, between adult learners using mindmapping versus traditional notetaking methods in a training lecture, and (2) there is no significant difference in posttest scores on an achievement test, administered 30 days after the course, between adult learners using mindmapping versus traditional notetaking methods in a training lecture. After a 1.5 hour instruction on mindmapping, the treatment group used mindmapping throughout the course. The control group used traditional notetaking. T-tests were used to determine if there were significant differences between mean posttest scores between the two groups. In addition, an attitudinal survey, brain hemisphere dominance survey, course dynamics observations, and course evaluations were used to investigate preference for mindmapping, its perceived effect on test performance, and the effectiveness of mindmapping instruction.^ This study's principal finding was that although the mindmapping group did not perform significantly higher on posttests administered immediately and 30 days after the course, than the traditional notetaking group, the mindmapping group did score higher on both posttests and reported higher ratings of the course on every evaluation criteria. Lower educated, right brain dominant learners reported a significantly positive learning experience. These results suggest that mindmapping enhances and reinforces the preconditions of learning. Recommendations for future study are provided. ^

Academic performance of college students in financial management and related prerequisite courses: A comparison of senior high school and commercial high school graduates attending Tamsui Oxford University College in Taiwan

In Taiwan, the college freshmen are recruited graduates of both senior high school and senior vocational school. The Ministry of Education (MOE) of the Republic of China prescribes the standards of curriculum and equipment for schools at all levels and categories. There exists a considerably different curriculum arrangement for senior high schools and vocational high schools in Taiwan at the present time. The present study used a causal-comparative research design to identify the influences of different post-secondary educational background on specialized course performance of college business majors. ^ The students involved in this study were limited to the students of four business-related departments at Tamsui Oxford University College in Taiwan. Students were assigned to comparison groups based on their post-secondary educational background as senior high school graduates and commercial high school graduates. The analysis of this study included a comparison of students' performance on lower level courses and a comparison of students' performance in financial management. The analysis also considered the relationship between the students' performance in financial management and its related prerequisite courses. The Kolb Learning Style Inventory (LSI) survey was administered to categorize subjects' learning styles and to compare the learning styles between the two groups in this study. The applied statistical methods included t-test, correlation, multiple regression, and Chi-square. ^ The findings of this study indicated that there were significant differences between the commercial high school graduates and the senior high school graduates on academic performances in specialized courses but not in general courses. There were no significant differences in learning styles between the two groups. These findings lead to the conclusion that business majors' academic performance in specialized courses were influenced by their post-secondary educational background. ^

Assessing the construct validity of a measure of learning agility

This study examined the construct validity of the Choices questionnaire that purported to support the theory of Learning Agility. Specifically, Learning Agility attempts to predict an individual's potential performance in new tasks. The construct validity will be measured by examining the convergent/discriminant validity of the Choices Questionnaire against a cognitive ability measure and two personality measures. The Choices Questionnaire did tap a construct that is unique to the cognitive ability and the personality measures, thus suggesting that this measure may have considerable value in personnel selection. This study also examined the relationship of this pew measure to job performance and job promotability. Results of this study found that the Choices Questionnaire predicted job performance and job promotability above and beyond cognitive ability and personality. Data from 107 law enforcement officers, along with two of their co-workers and a supervisor resulted in a correlation of .08 between Learning Agility and cognitive ability. Learning Agility correlated .07 with Learning Goal Orientation and. 17 with Performance Goal Orientation. Correlations with the Big Five Personality factors ranged from −.06 to. 13 with Conscientiousness and Openness to Experience, respectively. Learning Agility correlated .40 with supervisory ratings of job promotability and correlated .3 7 with supervisory ratings of overall job performance. Hierarchical regression analysis found incremental validity for Learning Agility over cognitive ability and the Big Five factors of personality for supervisory ratings of both promotability and overall job performance. A literature review was completed to integrate the Learning Agility construct into a nomological net of personnel selection research. Additionally, practical applications and future research directions are discussed. ^

Pavement performance prediction using adaptive logic networks

Pavement performance is one of the most important components of the pavement management system. Prediction of the future performance of a pavement section is important in programming maintenance and rehabilitation needs. Models for predicting pavement performance have been developed on the basis of traffic and age. The purpose of this research is to extend the use of a relatively new approach to performance prediction in pavement performance modeling using adaptive logic networks (ALN). Adaptive logic networks have recently emerged as an effective alternative to artificial neural networks for machine learning tasks. ^ The ALN predictive methodology is applicable to a wide variety of contexts including prediction of roughness based indices, composite rating indices and/or individual pavement distresses. The ALN program requires key information about a pavement section, including the current distress indexes, pavement age, climate region, traffic and other variables to predict yearly performance values into the future. ^ This research investigates the effect of different learning rates of the ALN in pavement performance modeling. It can be used at both the network and project level for predicting the long term performance of a road network. Results indicate that the ALN approach is well suited for pavement performance prediction modeling and shows a significant improvement over the results obtained from other artificial intelligence approaches. ^

Team Primacy Concept-based employee evaluation and job performance

This qualitative case study explored how employees learn from Team Primacy Concept (TPC)-based employee evaluation and how they apply the knowledge in their job performance. Kolb's experiential learning model (1974) served as a conceptual framework for the study to reveal the process of how employees learn from TPC evaluation, namely, how they experience, reflect, conceptualize and act on performance feedback. TPC based evaluation is a form of multirater evaluation that consists of three components: self-feedback, supervisor's feedback, and peer feedback. The distinctive characteristic of TPC based evaluation is the team evaluation component during which the employee's professional performance is discussed by one's peers in a face-to-face team setting, while other forms of multirater evaluation are usually conducted in a confidential and anonymous manner.^ Case study formed the methodological framework. The case was the Southeastern Virginia (SEVA) region of the Institute for Family Centered Services, and the participants were eight employees of the SEVA region. Findings showed that the evaluation process was anxiety producing for employees, especially the process of peer evaluation in a team setting. Preparation was found to be an important phase of TPC evaluation. Overall, the positive feedback delivered in a team setting made team members feel acknowledged. The study participants felt that honesty in providing feedback and openness to hearing challenges were significant prerequisites to the TPC evaluation process. Further, in the planning phase, employees strove to develop goals for themselves that were meaningful. Also, the catalyst for feedback implementation appeared to stem from one's accountability to self and to the client or community. Generally, the participants identified a number of performance improvement goals that they attained during their employment with IFCS, which were supported by their developmental plans.^ In conclusion, the study identified the process by which employees learned from TPC-based employee evaluation and the ways in which they used the knowledge to improve their job performance. Specifically, the study examined how participants felt and what they thought about TPC-based feedback, in what ways they reflected and made meaning of the feedback, and how they used the feedback to improve their job performance.^