The Covid-19 outbreak, which has been under the influence of Europe since then, continues to spread rapidly especially in the American continent. Looking at the current data, the virus has affected about 250 million people and has killed more than five million people. Especially with the rapid spread of the outbreak in the European continent, this issue started to be discussed in social media. In particular, Twitter is the most frequently used micro-blogging in this workspace. In this study, it is aimed to analyze the tweets shared by many people, organizations and government agencies through Twitter during the global COVID-19 outbreak with sentiment analysis using the VADER Sentiment Analysis method. The hashtags #covid19, #Covid, #pandemic, #social-distancing, #socialdistance, #covid-19, #corona-virius, #coronavirus, #Chinesevirus, #Chinese-virus were used in this study. With these hashtags, a total of 60,243,040 tweets were collected from Twitter between January 1, 2020 and July 1, 2020. In this study, we use the VADER to classify the sentiments expressed in Twitter data related to Covid-19 and the compound scores of the resulting tweets were divided into five categories: Highly Positive, Positive, Neutral, Negative, Highly Negative. In addition, in the study, the Wordcloud was used to visualize the most frequently collected text data monthly, and N-grams were applied to the tweets to better understand the content of the tweets. When the results obtained in the study are examined, the tweets shared about Covid-19 in different periods of the release reflect different sentimental situations.

With the rise of social media platforms, which have billions of users around the World, the dissemination of information has become easier than ever. The COVID-19 pandemic has increased the use of social media to discuss many topics, including vaccines. The aim of this study is to analyze public sentiment with Machine Learning of vaccine-related tweets obtained on Twitter in order to better understand the attitudes and concerns of social media users, especially regarding COVID-19 vaccines in Turkey. For this purpose, the majority voting method, which is an ensemble learning method, was developed by comparing the machine learning algorithm used in six different classification tasks and then via Support Vector Machine, XGBoost and Random Forest having the highest accuracy, in the study. Soft Voting method, which is one of the majority voting methods, has reached a success rate of 90.5%, with a higher success rate than both the Hard Voting approach and the other six individual machine learning approaches. With the Soft Voting method, which has the highest accuracy rate, 412,588 daily tweets from 153 days obtained from Twitter were analyzed and the results were reported. The findings of the study are very striking and differ from studies on other countries. As far as we know, this study is the first study to perform sentiment analysis on COVID-19 vaccines in Turkey, and it shows that it is a valuable and easily applied tool to monitor sensitivity to COVID-19 vaccines with sentiment analysis approach over social media.

Owning a house is one of the most important decisions that low and middle income people make in their lives. The real estate market is a significant factor of the national economy as much as it is important for individuals. Therefore, predicting real estate values or real estate valuation is beneficial and necessary not only for buyers, but also for real estate agents, economists and policy makers. This issue represents an active area of research, as individuals, companies and governments hold considerable assets in real estate. In this context, the aim of the study is to predict real estate prices with Machine Learning methods using the real estate sales data set in June and July 2021 belonging to the province of Ankara. In particular, it is to perform a comprehensive comparison on Machine Learning regression types methods that give successful prediction results in various but similar tasks, which are not included in the real estate literature. Real estate data obtained over the Internet was first included in a detailed data preprocessing process, and then Linear, Lasso and Ridge Regression, XGBoost and Artificial Neural Networks (ANN) methods were used on this dataset. According to empirical findings, XGBoost and ANNs appear as very important alternatives in predicting real estate sales prices.

İktisadi büyüme, bir ulus için tüm ekonomik faaliyetlerin nihai amacı olarak kabul edilmektedir. Bu önemi nedeniyle gerek ulusal politikaların başarısının değerlendirilmesinde gerekse uluslararası karşılaştırmalarda ilk akla gelen ekonomik gösterge durumundadır. Son dönemde çeşitli makine öğrenmesi yaklaşımları ve yapay sinir ağları modelleri de bu alanda kullanılmaya başlanmıştır. Bu çalışmanın amacı, literatürden farklı olarak karar ağacı modellemesini kullanarak iktisadi büyümenin kaynaklarını ortaya koymaktır. Bu bağlamda, literatürde büyümeye kaynaklık ettiği kabul gören tasarruf, sermaye birikimi, emek gücü, yenilik, kamu harcamaları, dış ticaret, beşeri sermaye, demografik unsurlar, sanayileşme, kurumsal faktörler gibi değişkenler analize dâhil edilmiştir. Çalışmanın bulguları, iktisadi büyüme için kurallar ve yüksek büyümenin sağlanmasına yönelik olarak politika yapıcılara referans olabilecek kanıtlar sunmaktadır. Çalışmanın bulguları, Türkiye Ekonomisinde büyümeyi teşvik etmede tasarruf, inovasyon, ihracat ve emeğin önemini ortaya koymaktadır. İktisadi büyüme için elde edilen kurallar, yüksek büyümenin sağlanmasına yönelik olarak politika yapıcılara referans olabilecek kanıtlar sunmaktadır.

For those who invest in real estate as an investment tool, as well as those who buy and sell real estate, the price of real estate should be predicted realistically and with the highest accuracy. It should be noted that the predict model should be the most appropriate representation of the underlying fundamentals of the market. Otherwise, the mistake to be made in the real estate valuation will cause some undesirable results such as inconsistent and unhealthy increase or decrease of the property tax, excessive gains or losses in favor of some groups, and adverse effects on investors and potential real estate owners. At this point, data-driven real estate valuation approaches are preferred more frequently to create highly accurate and unbiased estimates. However, the consistency, precision and accuracy of the models realized with machine learning approaches are directly related to the data quality. At this point, the effects of outlier detection on prediction performance in real estate valuation are investigated with a large data set obtained in this study. For this purpose, a heterogeneous data set with 70.771 real estate data and 283 variables, 4 different outlier detection methods were tested with 3 different machine learning approaches. The empirical findings reveal that the use of different outlier detection approaches increases the prediction performance in different ranges. With the best outlier detection approach, this performance increase was at a high 21,6% for Random Forest, with a 6,97% increase in average model performance.