The purpose of this project is to develop a model that is capable of recognizing daily basic human activities under real-world conditions, using data collected by a waist-mounted triaxial accelerometer and gyroscope built into a cellphone (in our study, a Samsung Galaxy S II). Activity recognition is formulated as a supervised classification problem, whose data is obtained via an experiment having 30 human subjects perform each of the activities. Our classification models have been trained...
This paper examines how people in Copenhagen interact with and perceive six of the main neighbourhoods in the capital. As a special case, an equivalent analysis is carried out for the popular food market Torvehallerne. The research was conducted by applying statistical analysis and natural language processing to captions from Instagram posts tagged with neighbourhood names. The analysis reveals that the Instagram segment have a positive perception of their neighbourhoods and like to displa...
Natural Language Processing is a flourishing aspect of Data Science in both academia and industry. This paper examines the technique of sentiment classification on text data using automated Machine learning approaches. It will delve into newly introduced embedding techniques and machine learning model pipelines. The models explored in this paper consist of a simple Naive Bayes classifier, as well as several Deep Neural Models, including a Gated Recurrent Unit (GRU) and a Long Term Short Te...
