Metadata augmented deep neural networks for wild animal classification

Camera trap imagery has become an invaluable asset in contemporary wildlife surveillance, enabling researchers to observe and investigate the behaviors of wild animals. While existing methods rely solely on image data for classification, this may not suffice in cases of suboptimal animal angles, lighting, or image quality. This study introduces a novel approach that enhances wild animal classification by combining specific metadata (temperature, location, time, etc) with image data. Using a dataset focused on the Norwegian climate, our models show an accuracy increase from 98.4% to 98.9% compared to existing methods. Notably, our approach also achieves high accuracy with metadata-only classification, highlighting its potential to reduce reliance on image quality. This work paves the way for integrated systems that advance wildlife classification technology.

GA-SVM for Evaluating Heroin Consumption Risk

There were over 70,000 drug overdose deaths in the USA in 2017. Almost half of those involved the use of Opioids such as Heroin. This research supports efforts to combat the Opioid Epidemic by further understanding factors that lead to Heroin consumption. Previous research has debated the cause of Heroin addiction, with some explaining the phenomenon as a transition from prescription Opioids, and others pointing to various psycho-social factors. This research used self-reported information about personality, demographics and drug consumption behavior to predict Heroin consumption. By applying a Support Vector Machine algorithm optimized with a Genetic Algorithm (GA-SVM Hybrid) to simultaneously identify predictive features and model parameters, this research produced several models that were more accurate in predicting Heroin use than those produced in previous studies. Although all factors had predictive power, these results showed that consumption of other drugs (both prescription and illicit) were stronger predictors of Heroin use than psycho-social factors. The use of prescription drugs as a strong predictor of Heroin use is an important though disturbing discovery but that can help combat Heroin use.