Development of Predictive Models for Rights Issues Using Financial Analysis Indices and Decision Tree Technique

This dissertation contributes with novel predictive modelling approaches to data-driven computational psychiatry and offers alternative analyses frameworks to the standard statistical analyses in psychiatric research. In particular, this document advances research in medical data mining, especially psychiatry, via two phases. In the first phase, this document promotes research by proposing synergistic machine learning and statistical approaches for detecting patterns and developing predictive models in clinical psychiatry data to classify diseases, predict treatment outcomes or improve treatment selections. In particular, these data-driven approaches are built upon several machine learning techniques whose predictive models have been pre-processed, trained, optimised, post-processed and tested in novel computationally intensive frameworks. In the second phase, this document advances research in medical data mining by proposing several novel extensions in the area of data classification by offering a novel decision tree algorithm, which we call PIDT, based on parameterised impurities and statistical pruning approaches toward building more accurate decision trees classifiers and developing new ensemblebased classification methods. In particular, the experimental results show that by building predictive models with the novel PIDT algorithm, these models primarily led to better performance regarding accuracy and tree size than those built with traditional decision trees. The contributions of the proposed dissertation can be summarised as follow. Firstly, several statistical and machine learning algorithms, plus techniques to improve these algorithms, are explored. Secondly, prediction modelling and pattern detection approaches for the first-episode psychosis associated with cannabis use are developed. Thirdly, a new computationally intensive machine learning framework for understanding the link between cannabis use and first-episode psychosis was introduced. Then, complementary and equally sophisticated prediction models for the first-episode psychosis associated with cannabis use were developed using artificial neural networks and deep learning within the proposed novel computationally intensive framework. Lastly, an efficient novel decision tree algorithm (PIDT) based on novel parameterised impurities and statistical pruning approaches is proposed and tested with several medical datasets. These contributions can be used to guide future theory, experiment, and treatment development in medical data mining, especially psychiatry.

To screen serum proteome biomarkers and establish predictive model with relation to the formation of portal vein tumor thrombi (PVTT) in hepatocellular carcinoma (HCC) patients. Serum samples were collected from 135 HCC patients, which were divided, into training set (including 33 HCC patients with PVTT and 62 HCC patients without PVTT) and blind testing set (including 18 HCC patients with PVTT and 22 HCC patients without PVTT). Special serum protein or peptide pattern was determined by SELDI-TOF-MS measurement after treating the sample onto WCX2 protein chip for each case. The obtained data were analyzed by BioMarker Wizard software to screen serum proteome biomarkers with relation to the formation of PVTT, while decision tree classification algorithm and blind validation were determined by Biomarker Patterns Software. Ranging from 1100 to 30 000 at the m/z value, 100 protein features were detected in the serum protein pattern stably. Among them, 6 protein peaks with the m/z value of 3478, 1314, 1744, 1725, 2022 and 3380 were upregulated, 10 proteins peaks with the m/z value of 8901, 9353, 9415, 8773, 2766, 2745, 8697, 7773, 8569 and 1373 were downregulated respectively in the group of HCC with PVTT. The 7 candidate protein peaks with the m/z value of 3478, 2022, 8901, 9415, 8773, 2766 and 2745 were selected to establish predictive model by BPS with a sensitivity of 75.8% (25/33) and specificity of 82.3% (51/62). An accuracy of 87.5% (35/40), sensitivity of 100% (18/18), specificity of 77.3% (17/22), positive predictive value of 78.3% (18/23), and negative predictive value of 100% (17/17) were validated in blind testing set. Sixteen candidate proteome biomarkers may be related with the formation of PVTT in HCC patients. Decision tree classification algorithm may have great clinical significance in predicting the formation of PVTT.

In general, obesity is calculated using the BMI, which is found to have limitations of use when we insert only one type of non-quantitative data as a crisp value, which is commonly used, or even create an obesity forecasting model using data mining techniques. If there is an input of weight and height data as a forecasting factor in various types of data, such as crisp value, estimated value and Fuzzy Linguistic Term value, etc. The algorithm will not possibly work under the data mining technique with such input data. This research presented the solution for this problem in order to enable data mining techniques under limitations of inserted data, including crisp value, estimated value and fuzzy linguistic term value, for forecasting obesity. The researchers created a fuzzy database system in the form of Conceptual Meta Schema to support fuzzy attributed data storage. In addition, there were 3 types of fuzzy attribute matching as follows; Type 1 was matching crisp value ​​with fuzzy linguistic term value, Type 2 was matching estimated value with fuzzy linguistic term value, and Type 3 was matching fuzzy linguistic term value with fuzzy linguistic term value, respectively. The attribute node from these fuzzy attributes would be selected to be used in the work of data mining techniques. Therefore, in this research, the researchers presented and compared the fuzzy model of working under the algorithm of data mining techniques in the form called fuzzy neural network and fuzzy decision tree techniques. The research result, it was found that the appropriate models for predicting obesity were fuzzy neural network algorithm with the neural network structure being 31-3-3, momentum at 0.2, learning rate at 0.3, under dividing data to test with cross-validation folds = 10 yielded accuracy value, precision value, recall value and f-measure at 84.3%, 82.7%, 84.3%, and 82.8%, respectively. Keywords: Obesity, Fuzzy Neural Net, Fuzzy Decision Tree, Fuzzy Attribute Matching, Forecasting Model

Human resource is one of the functions of a company that is considered as an asset. Therefo re, the theory of performance qualificat ion was adopted by the company in order to get an overview of employee performance. Furthermore, the company needs an effective method to predict the performance not only for the employees but also for the new applic ants. The goals of this research are to get a decision tree model of the employee performance. By learning employee data, the performance of the new applicants could be predicted. The study would provide the characteristic of new applicants who will give better performance than other applicants . The data from a company in Indonesia will have been used for this research. The data mining technique will be applied to the data of operators (such as admins, clerks, cashiers, machine operators, and security offi cers). The data mining technique was use d is decision tree. The decision tree technique was commonly used for a supervised learning data. The decision tree technique also has advantages compared others, because of its ability to produce information that is easy to understand. The result of this research shown the high dependency of employee performance with employment type (work contract). It also means that employees are encouraged to provide good performance to the company if those employees have become p ermanent employees. This research also showed that there is no relationship between employee performances with gender or position grade.

The leading treads towards social networking services had drawn massive public attention from last one and half decade. The amount of data that is uploaded to these social networking services is increasing day by day. So, their is massive requirement to study the highly dynamic behavior of users towards these services. This is a preliminary work to model the user patterns and to study the effectiveness of machine learning predictive modeling approaches on leading social networking service Facebook. We modeled the user comment patters, over the posts on Facebook Pages and predicted that how many comments a post is expected to receive in next H hrs. To automate the process, we developed a software prototype consisting of the crawler, Information extractor, information processor and knowledge discovery module. We used Neural Networks and Decision Trees, predictive modeling techniques on different data-set variants and evaluated them under Hits@10(custom measure), Area Under Curve, Evaluation Time and Mean Absolute error evaluation metrics. We concluded that the Decision trees performed better than the Neural Networks under light of all evaluation metrics.

Data mining refers to the application of data analysis techniques with the aim of extracting hidden knowledge from data by performing the tasks of pattern recognition and predictive modeling. This article describes the application of data mining techniques on educational data of a higher education institution in Croatia. Data used for the analysis are event logs downloaded from an e-learning environment of a real e-course. Data mining techniques applied for the research are cluster analysis and decision tree. The cluster analysis was performed by organizing collections of patterns into groups based on student behavior similarity in using course materials. Decision tree was the method of interest for generating a representation of decision-making that allowed defining classes of objects for the purpose of deeper analysis about how students learned.

The role of decision trees in software development effort estimation (SDEE) has received increased attention across several disciplines in recent years thanks to their power of predicting, their ease of use, and understanding. Furthermore, there are a large number of published studies that investigated the use of a decision tree (DT) techniques in SDEE. Nevertheless, in reviewing the literature, a systematic literature review (SLR) that assesses the evidence stated on DT techniques is still lacking. The main issues addressed in this paper have been divided into five parts: prediction accuracy, performance comparison, suitable conditions of prediction, the effect of the methods employed in association with DT techniques, and DT tools. To carry out this SLR, we performed an automatic search over five digital libraries for studies published between 1985 and 2019. In general, the results of this SLR revealed that most DT methods outperform many techniques and show an improvement in accuracy when combined with association rules (AR), fuzzy logic (FL), and bagging. Additionally, it has been observed a limited use of DT tools: it is therefore suggested for researchers to develop more DT tools to promote the industrial utilization of DT amongst professionals.

In this thesis a data-driven and linguistically interpretable intonation model for the automatic analysis and synthesis of fundamental frequency (F0) contours was developed. The intonation model: The model can be characterised as parametric, contour-based, and superpositional. F0 contours are treated as a superposition of global and local components. These components are anchored in a hierarchic prosodic structure defined by global and local segments which correspond roughly to intonation phrases and accent groups respectively. The stylisation of the F0 contours is carried out as follows: Within each global segment a linear F0 base contour is fitted. After the subtraction of this global baseline a third order polynomial is fitted to the F0 residual within each local segment. Subsequently, a symbolic description of the intonation inventory in form of global and local contour classes is derived by polynomial coefficient clustering. On the phonetic level, linear regression models adjust these abstract units to the respective prosodic context. As to the parametric and contour-based description, the model stands in the tradition of Fujisaki (1987), Mohler (1998b) and Taylor (2000). As to superposition, it stands in the tradition of Fujisaki (1987). As in Mohler und Conkie (1998) stylisation parameter clustering is carried out. Regarding the following aspects the approach chosen here provides additional benefit to intonation research: (1) The requirements for data preprocessing are comparably low. F0 stylisation was carried out in F0 sections at syllable nuclei, rendering an exact syllable segmentation unnecessary. The extraction of the prosodic structure is restricted to prosodic phrase boundaries guided by signal pauses, punctuation and partof-speech information. Pitch accent localisation and classification is not needed. Due to this a complete automation of the preprocessing steps with acceptable quality is achieved, so that there is no need for a manual data preparation by experts. This property allows for a fast adaptation of the model to new speech data and avoids inconsistencies caused by incomplete inter-labeller agreement. Due to the partly text-based definition of prosodic structure, automatic preprocessing includes a signal-text alignment needed for subsequent linguistic interpretation. (2) In contrast to the more complex stylisation functions of the models mentioned above, the polynomial stylisation chosen in this study guarantees an analytic approximation and thus a biunique relation between the F0 to be modelled and its abstraction. This property is essential to partition the F0 stylisations into intonation classes based on their contour similarity as well as for later linguistic interpretation. At the same time the chosen polynomial order is capable of capturing F0-coded prominence and boundary behaviour. Linguistic interpretation: The linguistic interpretability of local contour classes was examined for the concepts significance, informational novelty, and utterance finality. The approach chosen here can be described as follows: first, by automatic linguistic corpus analyses hypotheses about possible relations between contour classes and linguistic concepts are generated. These hypotheses are subsequently tested by perception experiments. By these means a systematic linguistic anchoring of the model was achieved in form of a decision tree to predict the linguistically appropriate contour class. The adequacy of its predictions was assured by a further perception test. Conclusion: It has been shown, that it is possible to build a perceptually acceptable and linguistically interpretable representation of intonation in a purely data-driven manner. This bottom-up approach guarantees consistency and easy adaptability of the model to new data. Due to its simultaneous signal proximity and linguistic anchoring, it covers the entire chain from text to signal and therefore can be used for intonation analysis and generation on a linguistic as well as on a phonetic-acoustic level. It is qualified for employment in speech technology applications as well as in phonetic fundamental research to automatically analyse raw speech data.

In order to survive in the modern world, organizations must be equipped with the mechanisms that not only maintain their competitive advantage, but also result in their progress and improvement. Prediction of banks’ performances is an important issue, and a poor performance in banks may primarily lead to their bankruptcy, thereby affecting national economics. The bank performance prediction model uses scientific and systematic approaches to diagnose the financial operations of institutes. According to a precise and strict evaluation, the model can detect the weakness of institutions in advance and provide early warning signals to related financial governments. In the present study, we have used three data mining models to predict the future performance of the banks accepted in Tehran Stock Exchange (TSE) and Iran Fara Bourse. Initially, 53 financial ratios were selected and, consequently, reduced to 28 using the fuzzy Delphi technique. The statistical population included 18 banks listed on TSE and Iran Fara Bourse, which provided their financial statements during the period of 2011 to 2017. Data were collected from the Codal site based on 28 financial ratios using C4.5 decision tree, AdaBoost, and Naive Bayes algorithm. According to the findings, the Naive Bayes algorithm was the optimal predictive model with the accuracy of 88.89%.

The developed credit score model represents an innovative and advanced way in the credit rating process that uses both financial and qualitative indicators in estimating creditworthiness. in order to develop the model a research has been carried out in the croatian banking sector and among the consulting companies involved in the process of creditworthiness. the model has been developed based on the findings and conclusions from the conducted research and contributes in the economic science, specifically in the field of innovation and competitiveness. The balance sheet data and the indicators originating from the financial statements can be used to determine how a company was performing in the past and how successful it was. But this information tells us very little about the quality of management, market share, company plans, production process, organizational structure, etc. and nothing about the directions in which the company is developing. This means that during the process of evaluation of creditworthiness a risk analyst needs much more information than the annual financial statements can provide. That is why qualitative factors must be also considered when estimating creditworthiness. The aim of this paper is to research the importance of the common use of the proposed qualitative and financial indicators in the process of estimating creditworthiness, and, in addition, to propose a model based on the resulting research findings. In the proposed credit score model, the final credit rating is determined by the synthesis and common scoring of financial and qualitative indicators, which ultimately reflects internal and external factors that directly affect the operations of any business entity. The proposed model will certainly allow a simpler and better rating score of small, medium and large companies, and with its simplicity it will contribute to the analysis of each business entity. In the model itself, the optimal choice of financial and qualitative indicators and their weighted values were used to make a more reliable credit rating score considering the internal and external factors affecting the business. Determining credit rating by applying financial and qualitative indicators provides comprehensive and credible information about the business entity and its business risks and thus allows the making of punctual and correct business decisions, i.e. wrong business decisions are prevented.

Air traffic controllers (ATCos) monitor flight operations and resolve predicted aircraft conflicts to ensure safe flights, making them one of the essential human operators in air traffic control systems. Researchers have been studying ATCos with human subjective approaches to understand their tasks and air traffic managing processes. As a result, models were developed to predict ATCo actions. However, there is a gap between our knowledge and the real-world. The developed models have never been validated against the real-world, which creates uncertainties in our understanding of how ATCos detect and resolve predicted aircraft conflicts. Moreover, we do not know how information from air traffic control systems affects their actions. This Ph.D. dissertation work introduces methods to evaluate existing ATCo action prediction models. It develops a prediction model based on flight contextual information (information describing flight operations) to explain the relationship between ATCo actions and information. Unlike conventional approaches, this work takes data-driven approaches that collect large-scale flight tracking data. From the collected real-world data, ATCo actions and corresponding predicted aircraft conflicts were identified by developed algorithms. Comparison methods were developed to measure both qualitative and quantitative differences between solutions from the existing prediction models and ATCo actions on the same aircraft conflicts. The collected data is further utilized to develop an ATCo action prediction model. A hierarchical structure found from analyzing the collected ATCo actions was applied to build a structure for the model. The flight contextual information generated from the collected data was used to predict the actions. Results from this work found that the collected ATCo actions do not show any preferences on the methods to resolve aircraft conflicts. Results found that the evaluated existing prediction model does not reflect the real-world. Also, a large portion of the real conflicts was to be solved by the model both physically and operationally. Lastly, the developed prediction model showed a clear relationship between ATCo actions and applied flight contextual information. These results suggest the following takeaways. First, human actions can be identified from closed-loop data. It could be an alternative approach to collect human subjective data. Second, the importance of evaluating models before implications. Third, potentials to utilize the flight contextual information to conduct high-end prediction models.

Background: Adolescence is a pivotal developmental period in the lifecycle. Aspects of shame, narcissism and social rank have all been associated with distress and wellbeing in adolescence, however no studies to date have investigated the impact of those constructs together in terms of their predictive value. Extant research has identified self-compassion as protective and associated with increased wellbeing. Aims: This study aimed to explore the relationships between external, internal and shame proneness; grandiose and vulnerable narcissism; social comparison and submissive behaviour, and self-compassion; and to explore the impact of those variables and the relationships between them upon psychological distress and wellbeing. Method: From the pragmatist approach this study adopted a cross-sectional, quantitative approach. Mid-adolescents aged 16-17 (N=142) were recruited and invited to complete a battery of self-report questionnaires via school or online survey. Correlation and regression analyses were performed to explore relationships and predictive associations between the variables and moderation analysis was performed to test the effects of self-compassion on the regression models. Results: This population was found to be low in wellbeing and moderate in distress. Multiple regression analyses found distress was predicted by internal shame, shame proneness, submissive behaviour and vulnerable narcissism; wellbeing was predicted by shame proneness; submissive behaviour and vulnerable narcissism (negatively) and social comparison and grandiose narcissism (positively). Self-compassion did not correlate with or predict distress, however its predictive power on wellbeing was marked. Moderation analyses showed self-compassion does not moderate distress or wellbeing. Conclusion: These novel findings suggest two predictive models for shame, narcissism and social rank in distress and wellbeing for an adolescent population, and show the impact of self-compassion. A more selfcompassionate attitude was related to increased wellbeing, hence those high in distress could benefit from compassion-based interventions and educational initiatives. Further investigations are warranted.

Based on the perspective of pig population system prediction, and accorded to principle of pig months transfer, this paper refers to the modeling principle and method of discrete population quantity prediction model. Then the prediction model of pork supply is derived and established: Firstly, the recursive model of pig population system and estimation model of pork supply was established. Then this study estimated the sum of monthly mortality and culling rate of breeding sows. Furthermore, the method for new left gilts in each month and estimation of breeding sows at each month of age was established. Last, this research established the estimation method model of the initial state of pig population. On this basis, an example calculation is made to predict the monthly pork supply in Heilongjiang Province from January 2016 to March 2018 in the future. The results showed that the prediction model of pork supply based on the prediction of pig population system is an effective perspective to study the forecast of pork supply. In the prediction stage, the prediction accuracy of the number of slaughtered fattened hogs was 96.36% and 97.54%, and the prediction accuracy of pork supply was 98.08% and 93.82%.This study not only lay a theoretical foundation for further study on the balance between pork supply and demand, but also helps to guide pork producers and governments at all levels to make relevant production decisions and plans. Keywords: pig population, breeding sow, pig production, pork supply prediction, population prediction model DOI: 10.25165/j.ijabe.20201302.5303 Citation: Zhang F, Wang F L. Prediction of pork supply via the calculation of pig population based on population prediction model. Int J Agric & Biol Eng, 2020; 13(2): 208–217.

The application of classification models in the credit rating of banking customers has been investigated in the present paper. Credit rating is one of the main applications of data mining in the banking industry. The customers' creditworthiness can be evaluated through credit ratings. The data related to banking customers is very huge, and various classification techniques can be used to explore the hidden pattern and knowledge in data set through data mining. Several studies have been performed on the use of data mining and classification techniques in the credit rating of banking customers. After preparing and preprocessing the data using the C5 decision tree algorithm in this paper, the classification model has been constructed and the credit rating of banking customers has been performed. A set of credit rating data has been used in this regard for teaching and testing the model. The results show that the developed model ranks banking customers with high accuracy by using decision tree making classification algorithms. The proposed classification model can also be used for a credit rating of new banking customers.

With the rapid development of digital economy, enterprises need to make financial management activities and development decisions based on a large number of financial data and information. This also exposes the shortcomings of the traditional financial analysis and processing mode in adaptability, functionality and efficiency. In this regard, based on the actual needs of enterprises, this paper will put forward a set of construction scheme of enterprise financial data visualization analysis system, give full play to the practical application advantages of data mining technology in the era of big data, change the original processing methods, and provide an efficient, reliable and convenient new way for enterprise financial data analysis. The whole system adopts B/S architecture design, the client design is based on VUE framework, and the server construction is based on Django framework, and the financial data is analyzed, mined and visually displayed by using Numpy, Pandas, Matplotlib and other class libraries in Python environment. Practice has proved that the system can support users to call K-means, Decision Tree, Logistic and other algorithm models online for multi-dimensional analysis of financial data, which realizes an important supplement to traditional financial analysis, and achieves the purpose of improving financial management efficiency and assisting enterprise development decision-making.