Sadi Evren Seker

Calculation of surface settlements caused by EPBM tunneling using artificial neural network, SVM, and Gaussian processes

Increasing demand on infrastructures increases attention to shallow soft ground tunneling methods in urbanized areas. Especially in metro tunnel excavations, due to their large diameters, it is important to control the surface settlements observed before and after excavation, which may cause damage to surface structures. In order to solve this problem, earth pressure balance machines (EPBM) and slurry balance machines have been widely used throughout the world. There are numerous empirical, analytical, and numerical analysis methods that can be used to predict surface settlements. But substantially fewer approaches have been developed for artificial neural network-based prediction methods especially in EPBM tunneling. In this study, 18 different parameters have been collected by municipal authorities from field studies pertaining to EPBM operation factors, tunnel geometric properties, and ground properties. The data source has a preprocess phase for the selection of the most effective parameters for surface settlement prediction. This paper focuses on surface settlement prediction using three different methods: artificial neural network (ANN), support vector machines (SVM), and Gaussian processes (GP). The success of the study has decreased the error rate to 13, 12.8, and 9, respectively, which is relatively better than contemporary research.

Estimation of Elastic Modulus of Intact Rocks by Artificial Neural Network

The modulus of elasticity of intact rock (Ei) is an important rock property that is used as an input parameter in the design stage of engineering projects such as dams, slopes, foundations, tunnel constructions and mining excavations. However, it is sometimes difficult to determine the modulus of elasticity in laboratory tests because high-quality cores are required. For this reason, various methods for predicting Ei have been popular research topics in recently published literature. In this study, the relationships between the uniaxial compressive strength, unit weight (γ) and Ei for different types of rocks were analyzed, employing an artificial neural network and 195 data obtained from laboratory tests carried out on cores obtained from drilling holes within the area of three metro lines in Istanbul, Turkey. Software was developed in Java language using Weka class libraries for the study. To determine the prediction capacity of the proposed technique, the root-mean-square error and the root relative squared error indices were calculated as 0.191 and 92.587, respectively. Both coefficients indicate that the prediction capacity of the study is high for practical use.

Ensemble classification over stock market time series and economy news

Aim of this study is applying the ensemble classification methods over the stock market closing values, which can be assumed as time series and finding out the relation between the economy news. In order to keep the study back ground clear, the majority voting method has been applied over the three classification algorithms, which are the k-nearest neighborhood, support vector machine and the C4.5 tree. The results gathered from two different feature extraction methods are correlated with majority voting meta classifier (ensemble method) which is running over three classifiers. The results show the success rates are increased after the ensemble at least 2 to 3 percent success rate.

Computerized Argument Delphi Technique

The aim of this study is the computerization of the argument Delphi method. The Delphi method is mainly designed for qualitative prediction within a group of experts, where the experts make predictions and a facilitator controls these predictions until the experts end up with a level of consensus. Argument Delphi, as opposed to the classical Delphi model, is built on the contradictions of the ideas of the experts. Argument Delphi mainly focuses on a discussion topic and asks experts to create new arguments and criticize other arguments from other experts. After a certain level of contradiction, the method yields an amount of contradictory, criticized arguments and builds a decision over these antitheses, as in the Hegelian approach. This is the first time the argument Delphi method has been modeled in a graph of arguments and the problem of qualitative decision has been transferred into a graph problem using Delphi method. This paper is also the first time that argument aggregation and evaluation methods have been proposed. Moreover, the computerized version of argument Delphi is applied to real-world problems using crowd involvement through Facebook. The problem is defined as the prediction of petroleum prices for the end of year and more than 100 contributors from all around the world argued and criticized each other. This paper also discusses the findings of this case study.

Author attribution on streaming data

The concept of novel authors occurring in streaming data source, such as evolving social media, is an unaddressed problem up until now. Existing author attribution techniques deals with the datasets, where the total number of authors do not change in the training or the testing time of the classifiers. This study focuses on the question, “what happens if new authors are added into the system by time?”. Moreover in this study we are also dealing with the problems that some of the authors may not stay and may disappear by time or may reappear after a while. In this study stream mining approaches are proposed to solve the problem. The test scenarios are created over the existing IMDB62 data set, which is widely used by author attribution algorithms already. We used our own shuffling algorithms to create the effect of novel authors. Also before the stream mining, POS tagging approaches and the TF-IDF methods are applied for the feature extraction. And we have applied bi-tag approach where two consecutive tags are considered as a new feature in our approach. By the help of novel techniques, first time proposed in this paper, the success rate has been increased from 35% to 61% for the authorship attribution on streaming text data.

Recurring and Novel Class Detection Using Class-Based Ensemble for Evolving Data Stream

Streaming data is one of the attention receiving sources for concept-evolution studies. When a new class occurs in the data stream it can be considered as a new concept and so the concept-evolution. One attractive problem occurring in the concept-evolution studies is the recurring classes from our previous study. In data streams, a class can disappear and reappear after a while. Existing studies on data stream classification techniques either misclassify the recurring class or falsely identify the recurring classes as novel classes. Because of the misclassification or false novel classification, the error rates increases on those studies. In this paper we address the problem by defining a novel ensemble technique “class-based” ensemble which replaces the traditional “chunk-based” approach in order to detect the recurring classes. We discuss the details of two different approaches in class-based ensemble and explain and compare them in detail. Different than the previous studies in the field, we also prove the superiority of both “class-based” ensemble method over state-of-art techniques via empirical approach on a number of benchmark data sets including Web comments as text mining challenge.

Web Based Reputation Index of Turkish Universities

This paper attempts to develop an online reputation index of Turkish universities through their online impact and effectiveness. Using 16 different web based parameters and employing normalization process of the results, we have ranked websites of Turkish universities in terms of their web presence. This index is first attempt to determine the tools of reputation of Turkish academic websites and would be a basis for further studies to examine the relation between reputation and the online effectiveness of the universities.

A Novel String Distance Function Based on Most Frequent K Characters

This study aims to publish a novel similarity metric to increase the speed of comparison operations. Also the new metric is suitable for distance-based operations among strings. Most of the simple calculation methods, such as string length are fast to calculate but does not represent the string correctly. On the other hand the methods like keeping the histogram over all characters in the string are slower but good to represent the string characteristics in some areas, like natural language. We propose a new metric, easy to calculate and satisfactory for string comparison. Method is built on a hash function, which gets a string at any size and outputs the most frequent K characters with their frequencies. The outputs are open for comparison and our studies showed that the success rate is quite satisfactory for the text mining operations.

Online anomaly detection for multi-source VMware using a distributed streaming framework

Anomaly detection refers to the identification of patterns in a dataset that do not conform to expected patterns. Such non‐conformant patterns typically correspond to samples of interest and are assigned to different labels in different domains, such as outliers, anomalies, exceptions, and malware. A daunting challenge is to detect anomalies in rapid voluminous streams of data.

Correlation between the Economy News and Stock Market in Turkey

Depending on the market strength and structure, it is a known fact that there is a correlation between the stock market values and the content in newspapers. The correlation increases in weak and speculative markets, while they never get reduced to zero in the strongest markets. This research focuses on the correlation between the economic news published in a highly circulating newspaper in Turkey and the stock market closing values in Turkey. In the research several feature extraction methodologies are implemented on both of the data sources, which are the stock market values and economic news. Since the economic news is in natural language format, the text mining technique, term frequency-inverse document frequency is implemented. On the other hand, the time series analysis methods like random walk, Bollinger band, moving average or difference are applied over the stock market values. After the feature extraction step, the classification methods are built on the well-known classifiers support vector machine, k-nearest neighborhood and decision tree. Moreover, an ensemble classifier based on majority voting is implemented on top of these classifiers. The success rates show that the results are satisfactory to claim the methods implemented in this study can be spread to future research with similar data sets from other countries.