Browse Articles

Based on the success of deep learning, recent works have attempted to develop a waste classification model using deep neural networks. This work presents federated learning (FL) for a solution, as it allows participants to aid in training the model using their own data. Results showed that with less clients, having a higher participation ratio resulted in less accuracy degradation by the data heterogeneity.

Despite the prevalence of PD, diagnosing PD is expensive, requires specialized testing, and is often inaccurate. Moreover, diagnosis is often made late in the disease course when treatments are less effective. Using existing voice data from patients with PD and healthy controls, the authors created and trained two different algorithms: one using logistic regression and another employing an artificial neural network (ANN).

Using facial recognition as a use-case scenario, we attempt to identify sources of bias in a model developed using transfer learning. To achieve this task, we developed a model based on a pre-trained facial recognition model, and scrutinized the accuracy of the model’s image classification against factors such as age, gender, and race to observe whether or not the model performed better on some demographic groups than others. By identifying the bias and finding potential sources of bias, his work contributes a unique technical perspective from the view of a small scale developer to emerging discussions of accountability and transparency in AI.

High school and university students have various motivations for participating in research, ranging from strengthening their applications for university to building skills for a research career. Jubair and Islam survey Bangladeshi high school and university students to uncover their motivations and inspirations for participating in research.

The authors develop a method for detecting fake AI-generated images from real images.

The authors looked at using genetic algorithms to look at the Indian labor market and what features might best explain any variation seen. They found that features such as economic growth and household consumption, among others, best explained variation.

The authors test the ability of planarians to remember conditioned stimuli following regeneration.

In this article the authors created an interaction map of proteins involved in colorectal cancer to look for driver vs. non-driver genes. That is they wanted to see if they could determine what genes are more likely to drive the development and progression in colorectal cancer and which are present in altered states but not necessarily driving disease progression.

Enzyme chemotaxis is a thermodynamic phenomenon in which enzymes move along a substrate concentration gradient towards regions with higher substrate concentrations and can be used to steer nanovehicles towards targets along natural substrate concentrations. In patients with Alzheimer’s disease, a gradient of tau protein forms in the bloodstream. Tau protein is a substrate of the enzyme CDK5, which catalyzes the phosphorylation of tau protein and can travel using chemotaxis along tau protein gradients to increasing concentrations of tau and amyloid-beta proteins. The authors hypothesized that CDK5 would be able to overcome these barriers of Brownian motion and developed a quantitative model using Michaelis-Menten kinetics to define the necessary parameters to confirm and characterize CDK5’s chemotactic behavior to establish its utility in drug delivery and other applications.

Machine learning algorithms are becoming increasingly popular for data crunching across a vast area of scientific disciplines. Here, the authors compare two machine learning algorithms with respect to accuracy and user-friendliness and find that random forest algorithms outperform logistic regression when applied to the same dataset.