Framework of Intelligent System for Machine Learning Algorithm Selection in Social Sciences

The ability to generate data has never been as powerful as today when three quintile bytes of data are generated daily. In the field of machine learning, a large number of algorithms have been developed, which can be used for intelligent data analysis and to solve prediction and descriptive problems in different domains. Developed algorithms have different effects on different problems.If one algorithmworks better on one dataset,the same algorithm may work worse on another data set. The reason is that each dataset has different features in terms of local and global characteristics. It is therefore imperative to know intrinsic algorithms behavior on different types of datasets andchoose the right algorithm for the problem solving. To address this problem, this papergives scientific contribution in meta learning field by proposing framework for identifying the specific characteristics of datasets in two domains of social sciences:education and business and develops meta models based on: ranking algorithms, calculating correlation of ranks, developing a multi-criteria model, two-component index and prediction based on machine learning algorithms. Each of the meta models serve as the basis for the development of intelligent system version. Application of such framework should include a comparative analysis of a large number of machine learning algorithms on a large number of datasetsfromsocial sciences.

Parametric Optimization of Integrated Circuit Assembly Process: An Evolutionary Computing-Based Approach

Strict demands for very tight tolerances and increasing complexity in the semiconductors’ assembly impose a need for an accurate parametric design that deals with multiple conflicting requirements. This paper presents application of the advanced optimization methodology, based on evolutionary algorithms (EAs), on two studies addressing parametric optimization of the wire bonding process in the semiconductors’ assembly. The methodology involves statistical pre-processing of the experimental data, followed by an accurate process modeling by artificial neural networks (ANNs). Using the neural model, the process parameters are optimized by four metaheuristics: the two most commonly used algorithms – genetic algorithm (GA) and simulated annealing (SA), and the two newly designed algorithms that have been rarely utilized in semiconductor assembly optimizations – teaching-learning based optimization (TLBO) and Jaya algorithm. The four algorithm performances in two wire bonding studies are benchmarked, considering the accuracy of the obtained solutions and the convergence rate. In addition, influence of the algorithm hyper-parameters on the algorithms effectiveness is rigorously discussed, and the directions for the algorithm selection and settings are suggested. The results from two studies clearly indicate superiority of the TLBO and Jaya algorithms over GA and SA, especially in terms of the solution accuracy and the built-in algorithm robustness. Furthermore, the proposed evolutionary computing-based optimization methodology significantly outperforms the four frequently used methods from the literature, explicitly demonstrating effectiveness and accuracy in locating global optimum for delicate optimization problems.

An ontology-based approach to the analysis of the acid-base state of patients at operative measures

Researchers working in various domains are focusing on extracting information from data sets by data mining techniques. However, data mining is a complicated task, including multiple complex processes, so that it is unfriendly to non-computer researchers. Due to the lack of experience, they cannot design suitable workflows that lead to satisfactory results. This article proposes an ontology-based approach to help users choose appropriate data mining techniques for analyzing domain data. By merging with domain ontology and extracting the corresponding sub-ontology based on the task requirements, an ontology oriented to a specific domain is generated that can be used for algorithm selection. Users can query for suitable algorithms according to the current data characteristics and task requirements step by step. We build a workflow to analyze the Acid-Base State of patients at operative measures based on the proposed approach and obtain appropriate conclusions.

Timetabling problem solving based on best-nests cuckoo search

The university courses timetabling problem (UCTP) is a popular subject among institutions and academics because occurs every academic year. In general, UCTP is the distribution of events through slots time for each room based on the list of constraints for instance (hard constraint and soft constraint) supplied in one semester, intending to avoid conflicts in such assignments. Under no circumstances should hard constraints be broken while attempting to fulfill as many soft constraints as feasible. this article presented a modified best-nests cuckoo search (BNCS) algorithm depend on the base cuckoo search (CS) algorithm. BNSC algorithm was achieved by dividing the nests into two groups (best-nests and normal-nests). The BNCS algorithm selection was limited to the best-nests to generate new solutions. The comparison between BNCS and basic CS based on the experimental result is achieved. For performance evaluation, the BNCS has been tested on four variant-size datasets. It was observed that the BNCS has performed high performance and is faster at finding a solution from CS.

Transformational machine learning: Learning how to learn from many related scientific problems

Almost all machine learning (ML) is based on representing examples using intrinsic features. When there are multiple related ML problems (tasks), it is possible to transform these features into extrinsic features by first training ML models on other tasks and letting them each make predictions for each example of the new task, yielding a novel representation. We call this transformational ML (TML). TML is very closely related to, and synergistic with, transfer learning, multitask learning, and stacking. TML is applicable to improving any nonlinear ML method. We tested TML using the most important classes of nonlinear ML: random forests, gradient boosting machines, support vector machines, k-nearest neighbors, and neural networks. To ensure the generality and robustness of the evaluation, we utilized thousands of ML problems from three scientific domains: drug design, predicting gene expression, and ML algorithm selection. We found that TML significantly improved the predictive performance of all the ML methods in all the domains (4 to 50% average improvements) and that TML features generally outperformed intrinsic features. Use of TML also enhances scientific understanding through explainable ML. In drug design, we found that TML provided insight into drug target specificity, the relationships between drugs, and the relationships between target proteins. TML leads to an ecosystem-based approach to ML, where new tasks, examples, predictions, and so on synergistically interact to improve performance. To contribute to this ecosystem, all our data, code, and our ∼50,000 ML models have been fully annotated with metadata, linked, and openly published using Findability, Accessibility, Interoperability, and Reusability principles (∼100 Gbytes).