There is a scarcity of fresh water in many rural communities where solar stills can be used to produce drinking water at a minimal cost. These stills use solar energy, which is a sustainable form of energy, and hence this can contribute towards achievement of United Nations (UN) Sustainable Development Goals (SDG). This study aims to develop empirical models of a solar stills based on experimental data obtained at Werrington South, New South Wales, Australia. Two solar stills were used in the experiment, a conventional design (Con-Still) and a con-still modified with adding extra thermal mass inside the still (mod-still). Regression analysis was adopted to develop prediction equations using Pi (productivity in L/m2/day) as the response variable and ambient temperature (Ta), sky temperature (Ts19), global radiation (Gh), and wind velocity (W) as the predictor variables. The mean and median productivity values of the mod-still were found to be 17%, and 22% higher than that those for the con-still. The proposed mod-still can be further improved and used in rural areas to produce fresh water from sea water and other forms of contaminated water.
Deception detection is regarded as a concern for everyone in their daily lives and affects social interactions. The human face is a rich source of data that offers trustworthy markers of deception. The deception or lie detection systems are non-intrusive, cost-effective, and mobile by identifying facial expressions. Over the last decade, numerous studies have been conducted on deception detection using several advanced techniques. Researchers have focused their attention on inventing more effective and efficient solutions for the detection of deception. So, it could be challenging to spot trends, practical approaches, gaps, and chances for contribution. However, there are still a lot of opportunities for innovative deception detection methods. Therefore, we used a variety of machine learning (ML) and deep learning (DL) approaches to experiment with this work. This research aims to do the following: (i) review and analyze the current lie detection (LD) systems; (ii) create a dataset; (iii) use several ML and DL techniques to identify lying; and (iv) create a hybrid model known as LDNet. By combining layers from Vgg16 and DeneseNet121, LDNet was developed and offered the best accuracy (99.50%) of all the models. Our developed hybrid model is a great addition that significantly advances the study of LD. The findings from this research endeavor are expected to advance our understanding of the effectiveness of ML and DL techniques in LD. Furthermore, it has significant practical applications in diverse domains such as security, law enforcement, border control, organizations, and investigation cases where accurate lie detection is paramount.
In todays competitive landscape, businesses grapple with customer retention. Churn prediction models, although beneficial, often lack accuracy due to the reliance on a single data source. The intricate nature of human behavior and high dimensional customer data further complicate these efforts. To address these concerns, this paper proposes a multimodal fusion learning model for identifying customer churn risk levels in financial service providers. Our multimodal approach integrates customer sentiments financial literacy (FL) level, and financial behavioral data, enabling more accurate and bias-free churn prediction models. The proposed FL model utilizes a SMOGN COREG supervised model to gauge customer FL levels from their financial data. The baseline churn model applies an ensemble artificial neural network and oversampling techniques to predict churn propensity in high-dimensional financial data. We also incorporate a speech emotion recognition model employing a pre-trained CNN-VGG16 to recognize customer emotions based on pitch, energy, and tone. To integrate these diverse features while retaining unique insights, we introduced late and hybrid fusion techniques that complementary boost coordinated multimodal co learning. Robust metrics were utilized to evaluate the proposed multimodal fusion model and hence the approach validity, including mean average precision and macro-averaged F1 score. Our novel approach demonstrates a marked improvement in churn prediction, achieving a test accuracy of 91.2%, a Mean Average Precision (MAP) score of 66, and a Macro-Averaged F1 score of 54 through the proposed hybrid fusion learning technique compared with late fusion and baseline models. Furthermore, the analysis demonstrates a positive correlation between negative emotions, low FL scores, and high-risk customers.