This study evaluates potential methods of harnessing sources of energy in everyday city life to generate electricity. By utilizing existing, underutilized (or non-utilized) sources of energy for electricity generation, use of hydrocarbons for power can be reduced. Specifically, natural light, artificial light, wind, and artificial water sources were evaluated. A series of tests to measure the efficacy of a solar panel in natural city light, a solar panel under different forms of artificial light, a wind turbine at average natural wind speeds, a hydroelectric turbine in the flow path of a sink, and a hydroelectric turbine in the flow path of a shower were conducted. […]

The purpose of this systematic review was to review different types of mosquito repellents in past research studies and papers and compare the findings of each, in order to draw conclusions. Mosquitoes have posed a huge issue to humans’ safety for centuries, as they have the ability to transmit fatal diseases. Usually, chemical factory made repellents are used against mosquitoes, but many species of mosquitoes have begun building resistance towards them, and they can cause harm to the environment, leading to the research and use of alternatives, in the form of organic repellents. This overall situation has sparked debates in the field based on the effectiveness of organic repellents, and the pros and cons of using them, as opposed to creating stronger chemical formulas. The methods of this project entailed performing a systematic review of articles extracted from research databases. […]

In 2013, the Virginia Medical Center found that 12 million Americans are impacted by clinical diagnostic errors each year — mistakes that result in 40,000 to 80,000 annual fatalities. The purpose of this study was to address this issue through the implementation of convolutional neural network (CNN) and long-short term memory (LSTM) machine learning algorithms in python, trained for dementia classification via MRI and demographic data. The CNN is a computer vision model capable of 2-D feature extraction through kernel convolution and pooling operations, in tandem with interconnected vectors (dense layers), containing individual elements referred to as neurons, which provide class predictions. The LSTM is a recurrent neural network composed of independent computational cells designed to perform sequential data analysis. Each of these cells are defined by a cell state, to which information can be appended or removed via regulatory gates that utilize the sigmoid function and pointwise multiplication to make decisions. […]