Yüksek Lisans Tezleri
Permanent URI for this collectionhttps://hdl.handle.net/20.500.11779/1785
Browse
12 results
Search Results
Master Term Project Predicting Customer Perfection on Brands Functional Near-Infrared Spectroscopy Measurements(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Kemerci, Emre; Koç, UtkuCustomer perception on the brands have importance to give strategic decisions by marketing professionals. In classical ways, customer perception on brands are researched through conducting field surveys. Similarly, neuromarketing discipline have studies on customer behaviors, their perceptions, communication techniques etc. under the frame of decision-making process of human. In neuromarketing, functional near-infrared spectroscopy (fNIRS) is a technology used to measure oxy and deoxy hemoglobin concentration in the tissues in order to enable to analyze hemodynamic responses of the brain activities. In this study, a group of participants’ activations of prefrontal cortex so the hemodynamic responses that were collected against a set of stimuli, which is a brand logo and adjective associated with the brand is used as dataset. Measured hemodynamic response metrics are oxygenated hemoglobin (HbO), deoxygenated hemoglobin (HbR), total hemoglobin (HbT) and Oxygenation (Oxy) and the dataset includes 168 participants’ measurements for 30 stimuli. In addition, the information regarding the responses of the participants and common perception of stimuli (field study results for same stimuli) are also exists in dataset. The aim of the project is to predict through machine learning algorithms whether relation between brand and the relevant adjective is Positive, Negative or Neutral using these feature set. As methodology of this study, fNIRS measurements in the data is cleaned and Null values are handled, measurements are consolidated per participant and stimuli with two different method as feature creation and classification algorithms are used as supervised learning to predict brand perception. In conclusion, performance of support vector classifier and XGBoosting algorithms are become very low, slightly over 50% accuracy despite the optimization with different classifier parameters. Further studies are addressed as performing feature engineering studies with different options.Master Term Project Analyzing the Drivers of Customer Satisfaction Via Social Media(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Yücel, Kadir Kutlu; Koç, UtkuSocial media became a great influence force during the last decade. Active social media user population increased with the new generations. Thus, data started to accumulate in tremendous amounts. Data accumulated through social media offers an opportunity to reach valuable insights and support business decisions. The aim of this project is to understand the drivers of customer satisfaction by public sentiments on Twitter towards a financial institution. Data was extracted from the most popular microblogging platform Twitter and sentiment analysis was performed. The unstructured data was classified by their sentiments with a lexicon-based model and a machine learning based model. The outcome of this study showed machine learning based model successfully overcame the language specific problems and was able to make better predictions where lexicon-based model struggled. Further analysis was performed on the extreme daily average sentiment scores to match these days with prominent events. The results showed that the public sentiment on Twitter is driven by three main themes; complaints related to services, advertisement campaigns, and influencers’ impact.Master Term Project Prediction of Up and Down Signalsın Selected Blues Chip Stocks(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Yıldız, Mustafa; Koç, UtkuEfforts have been made to predict the direction in which equity stocks will move in the capital markets. In most of these studies, Technical Analysis and Fundamental Analysis based models have been used. For daily price estimations, macroeconomic variables or financial ratios of financial instruments are used. On the other hand trade book data are taken into consideration in intraday price estimates. In this study, equity market data analytics, which are created by Borsa İstanbul as a benchmark for intraday price signals, are used. These analytics are derived from trade and order book data. For 5 minute periods, intraday price and equity market data analytics data sets are created, and different algorithms are tried over these data sets. The study is carried out using one-week data of 4 selected blue chip stocks. The signals for increase is 1, for decreases is -1 and 0 for non-change signals. As a result of the study, the decision jungle algorithm is the most successful algorithm. In addition this, the lack of volatility and liquidity in the market have caused overfitting problems in ensemble algorithms. According to the multiclass decision jungle confusion matrix, the positive true results for 1 (or increase of the price) are promising. If an investors can just use the algorithm for the price increase, it will be meaningful. The true positive ratio of 1, 54.5%, is too high when it is compared with its false trues value for decrease (or -1), which is just 13.6%. The difference between true positive and false negative (54.5% - 13.6%) will be the earning ratio for the investor, if he/she decides to invest the price increase of Yapi Kredi stock with the decision jungle algorithm. Although it is stated that big data algorithms (machine learning techniques) can give the best results for the data, domain knowledge related to the data is still very important. As it is seen in the study, in order to overcome the problems of overfitting or bias that occur in other studies, it is necessary to obtain sufficient domain knowledge in consultation with the experts and practitioners of the subject. In addition, the increase in the studies on intraday trading, which is a shallow area in the literature, will provide better results in the studies conducted on price forecasts in the future. In the results of this study, parallel with the literature, it is revealed that there is difficulty in estimating the stock price movements.Master Term Project Flight Delay Prediction(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Kurt, Mustafa; Taş Küten, DuyguThis study aims to create a model to predict flight departure delays. Various factors might affect a flight delay, and thus different features might be selected as input to create a model concerning priorities and the power of control over the features for the party who makes the analysis. In this study, domestic commercial flights in the U.S. operated in August 2018 are studied. Besides, airplane, passenger boarding, and cargo data are combined with flight data to benefit from possible insights related to these factors. For predicting the flight delays, machine learning methods such as decision trees, random forest, bagging classifier, extra trees classifier, gradient boosting and xgboost classifier are used and results are analyzed. Further studies could be adding extra features such as data related to flight planning, personnel data, loading data, data about technical processes to prepare a plane to a flight to improve prediction capacity.Master Term Project Predicting Customer Satisfaction Via Structed and Unstructured Data Using Classification and Regression(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2019) Danışman, Efehan; Küçükaydın, HandeAccording to different studies, retaining existing customers is five or more times more costly than acquiring new ones. This study aim to understand what customers expect from an airline using machine techniques. Dataset is scraped from Skytrax’s Airline Quality website and consists of 65947 observations with 17 columns consisting of one free format column that includes customer review. In order to do predict whether a customer recommends an airline or not, we try to utilize classification and regression algorithms. In addition to insights, this study also aims to compare the performance of the models and viability of using only free text in order to predict customer satisfaction.Master Term Project Tractor Sales Forecast Using Machine Learning(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2018) Tunay, Yiğitcan; Özlük, ÖzgürThis study presents a machine learning model to forecast tractor sales using four years of number of tractor sales based on year, month, city, town, brand and model provided by Turkey Statistical Institute. Tractor sales can vary depending on many different factors. Therefore, it is a challenging task for any company to estimate number of tractor sales that will be sold next year. Having the ability to predict that accurately will contribute companies in many distinct ways. Foreseeing market trends, keeping pace with the competition, delivering the right product to the right customer at the right time, reducing inventory costs, better production planning and cash flow management are major advantages of accurate forecasting. Within the scope of this study, models were developed to predict tractor sales using different statistical and machine learning methods. In further steps of the study, meaningful variables can be added to the dataset in order to reach a better result. Also, market share can be estimated by using different simulation methods which take into consideration those variables.Master Term Project Credit Card Fraud Detection Analysis and Machine Learning Application(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2018) Meker, Tuğrul; Özlük, ÖzgürCredit card fraud transaction is a common term for theft and fraud action involving a payment card such as payment or credit card or debit card as a source of funds in transactions. With increased usage of POS channel or internet in recent years, the risks of credit card fraud have increased. Mostly, these illegal activities start with a compromise of data associated with the account number or important information that required to start the financial transaction. After, literature review and exploratory data analysis, machine learning algorithms are going to use to decide whether the transaction is fraud or not. Logistic regression, decision tree, Naive-Bayes, decision forest and linear SVC’s classifier algorithms are used in this study. With re-sampling choices (random-under, random-over sampling & SMOTE), these algorithms’ performances are compared. Logistic regression, decision tree, and random forest provide best results in terms of accuracy metrics. Grid-Search is applied to those three algorithms. Decision tree algorithm is chosen as the best algorithm for credit card fraud detection. Python 3.7 is used in this study.Master Term Project Second-Hand Car Price Estimation Using Machine Learning(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2018) Kütükde, Şule; Özlük, ÖzgürThe ones who think to sell their cars always think about their cars’ second-hand market worth, at first. Both for the sellers and the buyers, it is crucially important to estimate the car’s realistic worth, in order not to suffer a loss of money or time. In this research, arabam.com’s advertisement data is obtained with the help of web scraping technique, and later machine learning algorithms like Linear Regression, Decision Tree, Random Forest and Gradient Boosting are applied for collected advertisement data in order to estimate cars’ prices. In addition, some hyperparameter tuning is applied for robust estimation. The models’ performances are discussed, and some remarks offered for further researches.Master Term Project Trangling Weratedogs Twitter Data To Create Interesting and Trustworthy Explosatory/Predictive Anaylses and Visulation Using Different Machine Learning Algorithms(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2018) Arı, Esra; Çakar, TunaSocial media usage has rapidly grown in recent years and knowledge in these environments increased due to this expansion. Therefore, doing exploratory and predictive analysis from intensive data of social media became so popular. However, almost all of the large datasets obtained are uncleaned / raw data. Therefore, the assessing and cleaning of the data is at least as important as the exploratory and predictive analysis. The open source WeRateDogs twitter account tweets have been gathered, assessed, cleaned, analyzed and predicted for this thesis. As a result of the study, it was understood that the most important and most time-consuming part of the predictive data analysis is the data gathering and cleaning. As a result of this project, probability of dog’s breed whether retriever or not is predicted from the tweet’s text body. 24 points increase (%34 change) in accuracy values has been achieved by doing oversampling in the data sets which contain low event observation. At the same time, the decision tree, logistic regression and random forest algorithms are compared and it is shown that the random forest's model performance is better than the others. The algorithm works 13 points better than logistic regression, 21 points better than decision tree.Master Term Project Credit Risk Models Using Machine Learning Models(MEF Üniversitesi, Fen Bilimleri Enstitüsü, 2018) Akman, Özkan; Çakar, TunaCredit scoring is an important subject in financial institutions, mainly in banks. I want to examine some machine learning techniques to find out a model that performs good in predicting or classifying the loaner person a good credit or a bad one by evaluating his/her demographic features as marital status, wealth, job seniority, monthly income and expenses.