Neural networks are used throughout modern society to solve many problems commonly thought of as impossible for computers. Fountain and Rasmus designed a convolutional neural network and ran it with varying levels of training to see if consistent, accurate, and precise changes or patterns could be observed. They found that training introduced and strengthened patterns in the weights and visualizations, the patterns observed may not be consistent between all neural networks.

Semantic segmentation - labelling each pixel in an image to a specific class- models require large amounts of manually labeled and collected data to train.

In this study, the authors seek to improve a machine learning algorithm used for image classification: identifying male and female images. In addition to fine-tuning the classification model, they investigate how accuracy is affected by their changes (an important task when developing and updating algorithms). To determine accuracy, a set of images is used to train the model and then a separate set of images is used for validation. They found that the validation accuracy was close to the training accuracy. This study contributes to the expanding areas of machine learning and its applications to image identification.

Reinforcement learning (RL) is a form of machine learning that can be harnessed to develop artificial intelligence by exposing the intelligence to multiple generations of data. The study demonstrates how reply buffer reward mechanics can inform the creation of new pruning methods to improve RL efficiency.

Neural machine translation (NMT) is a software that uses neural network techniques to translate text from one language to another. However, one of the most famous NMT models—Google Translate—failed to give an accurate English translation of a famous Korean nursery rhyme, "Airplane" (비행기). The authors fine-tuned a pre-trained model first with a dataset from the lyrics domain, and then with a smaller dataset containing the rhythmical properties, to teach the model to translate rhythmically accurate lyrics. This stacked fine-tuning method resulted in an NMT model that could maintain the rhythmical characteristics of lyrics during translation while single fine-tuned models failed to do so.

Hemani et al. tackled the problem of rampant hospital waste by implementing staff training to help inform hospital workers about proper waste disposal. The authors observed a significant increase in proper waste disposal after the training, showing that simple strategies, such as in-person classroom training and posters, can have a profound effect on limiting improper waste handling.

Wearing face masks has become a common occurrence in everyday life and during athletics due to the spread of diseases. This study tested if masks would affect blood percent saturation of hemoglobin (SpO₂) during treadmill exercise. The data analysis showed that mask type, time, and the interaction of mask type and time were significant results, regardless of physical ability. These results may assist athletes in understanding the differences between training and competing with and without a mask.

In this work, the authors investigate the accuracy with which two different population growth models can predict population growth over time. They apply the Malthusian law or Logistic law to US population from 1951 until 2019. To assess how closely the growth model fits actual population data, a least-squared curve fit was applied and revealed that the Logistic law of population growth resulted in smaller sum of squared residuals. These findings are important for ensuring optimal population growth models are implemented to data as population forecasting affects a country's economic and social structure.

The authors develop a new method for training machine learning algorithms to differentiate between hate speech and cultural speech in online platforms.

The purpose of the study was to determine whether graph-based machine learning techniques, which have increased prevalence in the last few years, can accurately classify data into one of many clusters, while requiring less labeled training data and parameter tuning as opposed to traditional machine learning algorithms. The results determined that the accuracy of graph-based and traditional classification algorithms depends directly upon the number of features of each dataset, the number of classes in each dataset, and the amount of labeled training data used.